Introduction

Bullying is widely acknowledged in the literature as a group phenomenon (Garandeau et al., 2014; Saarento & Salmivalli, 2015; Salmivalli, 2010; Shin, 2022; Veenstra & Huitsing, 2021). From this perspective, bullying is defined as a relational process involving several young people who socially influence each other to perpetrate or exacerbate violence against victims who struggle to defend themselves (Pikas, 1975; Sutton & Smith, 1999). Given that schools bring together large numbers of teenagers, bullying often manifests within this environment. Bystanders assume a pivotal role in the emergence and the evolution of bullying, either by attributing significance to violence or, conversely, by opposing it and thereby reducing its impact (Saarento & Salmivalli, 2015; Salmivalli, 2010; Salmivalli et al., 1996; Van Rijsewijk et al., 2016). Social influence also impacts bystanders, influencing their responses to bullying (behavior), their evaluation of violence (attitudes), and the interpretation they ascribe to it (cognitions). Instances where bystanders mimic or exacerbate violence (behavior) often coincide with supportive attitudes and cognitions toward it. Conversely, bystanders less frequently challenge bullying. Research shows that only 17.3% of youths adopt prosocial attitudes, cognitions, or behaviors toward victims – such as intervening on behalf of the victim (physically or verbally), seeking retribution against the perpetrator, or offering emotional support to the victim (e.g., expressing empathy, providing comfort) – while 50% either exacerbate or imitate the violence (Eisenberg et al., 2010; Machackova et al., 2018; Salmivalli et al., 1996; Van Rijsewijk et al., 2016). Moreover, engaging in prosocial behaviors, attitudes, or cognitions toward victims presents challenges if such actions are not already valued among peers. This challenge is particularly pronounced during adolescence, given the identity development that adolescents navigate during this phase, which amplifies the significance of peers relationships and their desire to conform to group norms (Hernandez et al., 2014; Laursen & Veenstra, 2021; Ragelienė, 2016).

This research explores social influence among adolescent bystanders in bullying situations from a psycho-educational perspective. Specifically, given the pivotal role of peers during adolescence, this study focuses on social influence within adolescent relationships, encompassing both existing connections and those individuals aspire to cultivate (aspirational). Employing a mixed methodology, predominantly drawing on qualitative empirical data supplemented by quantitative insights, the research contributes to understanding various forms of social influence in bullying (conformity, compliance, and obedience) within peer relationships (group, privileged, and aspirational), as well as the interplay between these different forms. This study is part of a larger research project on school bullying led by Moody et al. (2019, 2020) and carried out in the canton of Valais in Switzerland, supported by the Minister Education of Valais. Its objectives include evaluating bullying prevalence in schools and elucidating the psychosocial trajectories of individuals involved in bullying.

Social Influence and Relationships

Social influence refers to the fact that young people affect and are affected in return by their peers’ behaviors, often causing their own behaviors, attitudes, or cognitions to change (Brown & Larson, 2009; Cialdini & Griskevicius, 2010; Giletta et al., 2021; Harkins et al., 2017; Laursen & Veenstra, 2021). Social influence is often reciprocal between young people (Brown & Larson, 2009). Although the strength of the links between behaviors, attitudes, and social cognitions is debated in the scientific literature (Howe & Krosnick, 2017), it is broadly agreed that changes in one may lead to changes in the other. Thus, social influence consists of a change in direction in young people’s attitudes, cognitions, or behaviors prompted by the ways of doing, thinking, and being of their peers or the group to which they belong. This change can positively and negatively affect teenagers (Giletta et al., 2021; Laursen & Veenstra, 2021).

According to the Influence-Compatibility model, social influence emerges between individuals when they actively seek to develop or maintain relationships between them by accentuating their reciprocal similarities (Giletta et al., 2021; Laninga-Wijnen et al., 2020; Laursen & Veenstra, 2021; Veenstra & Huitsing, 2021). This is the case during adolescence due to the identity development that young people face (Hernandez et al., 2014; Laursen & Veenstra, 2021; Ragelienė, 2016; Tajfel & Turner, 1979, 1986). The individual development of young people takes place in particular through their relationships with each other, allowing them to compare and discover themselves. Such a process accentuates their search for mutual similarities and increases social influence at this stage of their life (Hernandez et al., 2014; Laursen & Veenstra, 2021; Ragelienė, 2016). Therefore, when confronted with a bullying situation, bystanders tend to adjust their attitudes, cognitions, and behaviors with those of young people with whom they have or wish to develop relationships. Research shows (Laursen & Veenstra, 2021; Peets et al., 2015; Saarento & Salmivalli, 2015; Stahel & Moody, 2023; Tajfel & Turner, 1986; Wang et al., 2022) that relationships between young people develop and are maintained according to a them/us logic, similarities in their own group are reinforced, at the expense of excluded or assaulted victims. Considering the role of social influence in interpersonal relations, this study focuses on the interplay between different forms of social influence and peer relationships in bullying.

Different Forms of Social Influence

Three main concepts underlie social influence (Harkins et al., 2017): (a) conformity, which reflects the fact that by observing peers’ behaviors, teenagers behave similarly and adapt their cognitions and attitudes in order not to be perceived as different, (b) compliance, which refers to behaviors, cognitions or attitudes adopted by young people in response to implicit or explicit requests from their peers, and (c) obedience, referring to behaviors, attitudes or cognitions that teenagers adopt as a result of implicit or explicit orders or requests from peers who have authority/power over them (Harkins et al., 2017; Laursen & Veenstra, 2021). Some research on bullying has provided insights into forms of social influence (Veenstra & Huitsing, 2021; Veenstra & Lodder, 2022), but systematic description and study of how they interact are lacking.

The search for conformity often happens through direct observation of peers’ behaviors. Some scholars have, however, suggested that conformity can take more indirect forms when it results from the interpretation of social norms in place among young people (Harkins et al., 2017; Laninga-Wijnen et al., 2020; Stahel & Moody, 2023; Veenstra & Lodder, 2022). In this case, young people indirectly perceive social norms defining what behaviors, attitudes, or cognitions are valued based on their understanding of their peers’ expectations. Teenagers then seek to conform to these social norms to be accepted by peers, according to the person-group dissimilarity model (Cook et al., 2010; Fandrem et al., 2010; Salmivalli, 2010; Wang et al., 2022). Since social norms vary depending on groups, which may be more or less favorable to bullying, the reactions of bystanders change accordingly (Laursen & Veenstra, 2021; Stahel & Moody, 2023; Veenstra & Lodder, 2022). This contradicts, to some extent, the way the literature rather permanently defines bullying roles (bystander, perpetrator, and victim) or merely combines them – e.g., the bully-victim, as documented by Olweus (1991).

Conformity may be more influential in bullying when it is combined with compliance. Cialdini and Griskevicius (2010) suggest that one can socially influence others through explicit requests. Teenagers may directly ask their peers to be, do, or think like them, thereby consolidating their similarities and further developing their relationships. Requests can also be indirect: By conforming to social norms, adolescents adjust to their peers’ expectations. Depending on their success, they may gain or lose social rewards (e.g., greater social recognition or acceptance, creating or sustaining relationships vs. being socially rejected or losing a relationship). The possibility of gaining or losing social rewards influences the reactions of bystanders, depending on whether it hinders or facilitates creating and sustaining peer relationships (Giletta et al., 2021; Laninga-Wijnen et al., 2020; Peets et al., 2015; Saarento & Salmivalli, 2015; Spadafora et al., 2020; Stahel & Moody, 2023; Veenstra & Lodder, 2022; Wang et al., 2022).

Recent research (Garandeau et al., 2014; Giletta et al., 2021; Peets et al., 2015; Saarento & Salmivalli, 2015; Veenstra & Huitsing, 2021; Veenstra & Lodder, 2022) shows that conformity and compliance are closely related to obedience; which suggest that all forms of social influence are interconnected. When studying bullying (Garandeau et al., 2014; Saarento & Salmivalli, 2015), obedience stems from how some adolescents adapt socially to others more salient or dominant within their peer group. It should be noted that dominance leads to perceived popularity among their peers. This form of popularity differs from sociometric popularity, which characterizes individuals who generate positive emotions or feelings in others (Brown & Larson, 2009; Kindelberger, 2018; Salmivalli, 2010).

In bullying, obedience may lead bystanders to adopt the same behaviors, attitudes, and cognitions as those of peers perceive as popular in the hope of gaining social acceptance or recognition, developing relationships with them, and ultimately protecting themselves from potential victimization (Saarento & Salmivalli, 2015; Veenstra & Lodder, 2022). In the face of peers perceived as popular, bystanders risk exclusion or other forms of bullying whenever they deviate from the social norms those shape. Such deviation generates social pressure on popular students, prompting them to act as perpetrators – consciously or unconsciously – to maintain existing social norms (Saarento & Salmivalli, 2015; Veenstra & Lodder, 2022). This trend remains prevalent even in environments such as classrooms, where multiple social norms can coexist – e.g., prosocial and violent (Laninga-Wijnen et al., 2020).

Different Forms of Relationships

As highlighted by Laursen and Veenstra (2021), adolescents may adapt their behaviors, attitudes, and cognitions even before entering relationships with peers, suggesting that the aspiration to be in a relationship already exerts social influence. Bukowski and Hoza (1989) distinguish two forms of relationships: (a) those that emerge when teenagers are accepted into one or more peer groups (e.g., peer affiliations; Brown & Larson, 2009) and (b) the creation of privileged relationships between at least two young people (Newcomb & Bagwell, 1995). These different forms of relationship give rise to various levels of analysis of social influence: privileged relationships refer to an interpersonal level of influence, while influence between members of a group to the intragroup level (Harkins et al., 2017; Mateeva & Dimitrov, 2013; Tajfel & Turner, 1979, 1986). To date, little research has examined links between the different levels of analysis derived from relationships and forms of social influence. Moreover, while perpetrators’ relationships with other peers and those of victims have been studied (Brown & Larson, 2009; Fandrem et al., 2010), how bystanders’ relationships and social influence are at play in bullying situations remains partly unexplored.

At the intragroup level, research on peer affiliation (Brown & Larson, 2009; Cook et al., 2010; Harkins et al., 2017; Solomontos-Kountouri & Strohmeier, 2021; Wang et al., 2022 ) considers the sense of belonging and identification within a peer group to understand the impact of social influence on individuals. This consideration is essential during adolescence since belonging to a group multiplies the opportunities to compare oneself with peers and, thus, to explore one’s identity (Hernandez et al., 2014). Moreover, behind bullying may lie a need to develop or maintain peer affiliations; when teenagers do not feel socially recognized or accepted, whether in or outside groups, engaging in violence may be a way of creating or sustaining peer affiliations (Saarento & Salmivalli, 2015; Solomontos‐Kountouri & Strohmeier, 2021). In addition, when bullying bystanders feel a sense of belonging, identify with, or aspire to be part of a peer group that values bullying, they are more likely to amplify, imitate, or engage in such aggressive behavior. Conversely, if the social norms within the group lean more towards prosocial behavior, bystanders are less likely to get involved in bullying (Laninga-Wijnen et al., 2020).

Unlike peer affiliations, privileged relationships are created between a few young people (at least two, including dyadic relationships). Privileged relationships can emerge within peer groups based on peer affiliation but do not necessarily equate to friendship. On an interpersonal level, it is crucial to consider the intensity of privileged relationships. Friendships appear gradually as teenagers become close to each other; their high intensity makes them unique (Meter & Card, 2016). In adolescence, this intensity develops not only through intimacy but also through self-disclosure. Moreover, identity exploration is of higher quality when it takes place in the context of friendships, as safe spaces where young people can exchange information about themselves with less fear that it will be used against them (Laursen & Veenstra, 2021; Ragelienė, 2016).

Not all friendships exert the same level of social influence on adolescents. Prosocial friendships tend to influence young people’s behaviors, attitudes, and cognitions weaklier than those that value violence (Laninga-Wijnen et al., 2020). Thus, if bystanders have prosocial friendships, they may reinforce or imitate violence toward victims. Young people tend to be more tolerant and respectful of their friends’ behaviors, attitudes, or cognitions, even if they differ from theirs (Laninga-Wijnen et al., 2020). Nonetheless, the lack of similarity can also become unbearable for friends, leading to their friendship breakdowns (Laursen & Veenstra, 2021).

Benefits of a Mixed Methodology

This study uses a mixed methodological device. Mixing methods is beneficial when studying interacting elements, such as the different forms of social influences and bystander relationships with peers in bullying (Moseholm & Fetters, 2017). First, data collected by one method can be complemented by another. Interactions can guide or even deepen the analysis of some dimensions from other perspectives. This can also compensate for the respective biases of each method. This study employs several methods, including focus groups, individual interviews, and an online questionnaire. While the survey offers a systematic view of bullying, it is limited by the number of variables at play. This limitation is counterbalanced by the density of the data obtained through qualitative methods (see also Hong & Espelage, 2012; Moseholm & Fetters, 2017; Schoonenboom & Johnson, 2017; Uprichard & Dawney, 2019). More specifically, data collected by qualitative methods (focus groups and individual interviews) form the basis for its analysis. This means that qualitative data is favored to describe and better understand the interplay between forms of social influence and peer relationships. Quantitative data – collected via online questionnaires – serve to support them, or in other words, to underpin them if necessary. This leads to qualify the mixed methodology mobilized in this study as “qualitative dominant” (Schoonenboom & Johnson, 2017, p.113).

By prioritizing qualitative data, a second benefit emerges from employing a mixed methodology: giving voice to young people on their own realities. It, therefore, becomes possible to include their point of view and understandings about the various forms of social influence between them and their relationships. Qualitative dominant research supports young people using their own words to express themselves. This produces a more dynamic understanding of phenomena concerning young people (Moody et al., 2021), bullying, and a more fluid vision of their ways of being, thinking, and doing (Hong & Espelage, 2012).

Third, inviting young people to express their viewpoints on the object of study also enables researchers to consider aspects that might have yet to be anticipated. This is particularly the case in this research, where different intensities in peer relationships are identified through the use of certain words by young people at the expense of others (Moody et al., 2020).

Finally, this mixed methodology considers the needs of participants and the sensitive nature of certain subjects. Given that this study considered that bullying roles (victims, perpetrators, and bystanders) may not be fixed, the voices of representatives of any of those roles were included. Victims and perpetrators were questioned through individual interviews to ensure confidentiality and to mitigate the risks of reprisal from their peers when talking about delicate situations. Bystanders were, on their part, involved in focus groups with other young people to encourage discussion, particularly among more introverted teenagers. Their risk of retaliation was mitigated by making it challenging to attribute blame to other participants for a discussion they all participated in. Parallelly, to enhance the chances of collecting a diverse range of bullying experiences, an online questionnaire was administered to youth, irrespective of their role in bullying. This questionnaire ensured the anonymity of participants by interviewing both those who have experienced bullying and those who have not, thereby minimizing the risk of being exposed to reprisals. It also enabled the gathering of experiences of participants who may be reluctant to participate in qualitative methods for fear of being made visible to their peers.

Current Study

There are at least three critical gaps in understanding social influence in relation to bullying. First, the different forms of social influence are not directly described and understood, nor are their interactions. Second, the forms of social influence are to be studied in relation to the various forms of peer relationships, including those they aspire to create with other individuals (aspirational). Third, there is a particular need to extend the study of these relationships to bystander students, recognizing their central role in the emergence and evolution of bullying rather than focusing on those between victims and perpetrators (Brown & Larson, 2009; Fandrem et al., 2010). To address these gaps, this research poses the following research question: How do different forms of social influence and peer relationships intersect and affect student bystanders of bullying? All aspects of this question are analyzed primarily based on qualitative data (focus groups and individual interviews). If needed, qualitative data are supported by quantitative ones (online questionnaires); for example, relative frequencies of young people’s behaviors, attitudes, and cognitions support the analysis of social influence, reasons for its emergence, and links of social influence with the relationships of bystanders.

Because of the need for similarities between teenagers, conformity is expected to be expressed in all relationships that bystander students have, whether on an intragroup or interpersonal level, including those they aspire to develop. More specifically, given the importance that young people place on belonging to and identifying with a peer group, it is expected that the relationships that bystanders aspire to create will have a strong social influence; this could give rise to an additional level of social influence analysis, the intergroup level.

While conformity occurs in most forms of relationships, it is assumed that compliance and obedience, which strongly interact with each other, only exist in certain relationships. These two forms of social influence are more related to recognition and social acceptance, which is why they supposedly mainly appear in group relationships.

Additionally, it is expected that when bystander students have privileged relationships with other actors in bullying situations – whether there are victims or perpetrators – these relationships may regulate and mitigate other forms of social influences that appear in other relationships. The closer the bystander is to the victim or perpetrator, the more likely this regulation will occur. Bystanders who are friends with victims or perpetrators tend to take their side, even if it happens to defy their peers’ expectations.

Methodology

Participants and Selection Process

The application of a mixed methodology for data collection in this study, including focus groups, individual interviews, and an online questionnaire, was carried out with several sub-samples of participants, presented in Table 1. Intermediaries (principal, school mediators, psychologist, or adult of trust) supported the research team in contacting individuals of the various sub-samples for two ethical reasons. On the one hand, to respond to questions from participants on a sensitive topic and, on the other, to respect the complete anonymity of children and youth contacted before the research.

Table 1 Description of sub-samples
Full size table

Table 1 shows that 3,275 participants aged between 10 and 17 participated in this study. According to the methods of inquiry, they were divided into three sub-samples.

The first sub-sample consists of participants who were asked to complete an online questionnaire and were selected through their schools in two separate stages. The first stage involved a random draw from all compulsory secondary schools in Valais, based on predefined criteria such as regions (plain and side valleys), school size (i.e., number of classes), and linguistic characteristics (French or German), to have a representative sub-sample of the region studied. Once the schools had been selected, the second stage involved inviting all students in grades 9–10Footnote 1 (12–14 years; 302 classes in total) to complete the online questionnaire.

The second sub-sample comprises participants invited to participate in a focus group, drawn randomly, selected from compulsory secondary schools in Valais that were not involved in sub-sample 1 (online questionnaire). As sub-samples were not surveyed simultaneously, selecting students through different schools minimized the risk of sharing information about questions asked during data collection. Once the schools had been selected, the next step was identifying students who self-identified as bystanders to bullying. To do this, school mediators presented the research to all grades 9 and 10 students. If they had witnessed a bullying situation, they were encouraged to announce themselves to participate in the research project via a short form, noting that all students had to return this form to the mediator.

The remaining sub-sample (3) comprised participants asked to participate in an individual interview because they self-identified as victims or perpetrators. Intermediaries supported selecting students willing to share their bullying experiences during an interview. Since discussing bullying could evoke painful emotions, participants who wished could ask their psychologist to participate in their exchange to provide support, which three participants requested.

Approach, Procedure, and Instrument

This research uses a comprehensive and iterative approach, complementing the qualitative dominant mixed methodology. It is comprehensive because the object of study – the interplay between different forms of social influence and relations between bystanders and other peers in bullying situations – is progressively constructed from the data collected and, thus, from the meaning that the participants give to them (Charmillot & Seferdjeli, 2002). It is iterative, unfolding in a constant back-and-forth movement between the data collected and their analysis (Moseholm & Fetters, 2017). Quantitative and qualitative methods were applied sequentially in 2019 (Schoonenboom & Johnson, 2017): Online questionnaires were administered from January to March, followed by nine focus groups and seven individual interviews between April and July. It is important to note that participants took part in only one data collection method, as their participation in the research is anonymous, and it could have been relatively complex to identify the words collected through other means.

The European Observatory of School Violence first developed and validated the online questionnaire. Initially, this online self-report instrument was designed to examine “the relationship between victimization and components of school climate such as general school appreciation of the school, relationships between the members of the school community, and quality of education” (Konstantina & Pilios-Dimitris, 2010, p.96). It has been widely used in many previous European empirical studies, and its psychometric reliability has been solidly confirmed in the scientific literature (see notably Debarbieux, 1998; Debarbieux, 2004; Debarbieux, 2011; Konstantina & Pilios-Dimitris, 2010; Debarbieux, 2018; Debarbieux & Moignard, 2022). To meet the needs of this study, the research team adapted the questionnaire to the Swiss context.

This online questionnaire consisted of 67 closed and mixed questions, the latter offering the possibility to provide details for the “other” response mode. In line with the questionnaire’s initial objective, questions are grouped into the following 11 topics: school climate, verbal bullying, physical bullying, sexual bullying, cyberbullying, appropriation violence, group dynamics, coping strategies, school dropout, socioeconomic status, and respondents’ identity. Some questions are subdivided into sub-questions if participants’ answers require clarification.

The research team included the theme of group dynamics when adapting the questionnaire to the Swiss context and to this study. The research team formulated questions and sub-questions based on theoretical sensitivity rather than pre-established theories (cf. Table 2, Charmillot & Seferdjeli, 2002). Thus, in line with the challenges of a comprehensive and iterative approach (Charmillot & Seferdjeli, 2002; Moseholm & Fetters, 2017), it was only when the collected data were analyzed that theoretical knowledge was combined, enabling the questions and sub-questions of the online questionnaire to be linked to specific concepts related to social influence and peer relationships (cf. column 1 in Table 2). It should be noted that although this study focuses on a restricted set of questions and sub-questions, all items queried via the online questionnaire were analyzed as part of the larger research project to which it belongs (see Moody et al., 2019; 2020).

The online questionnaire was tested on same-aged volunteers before administering it to ensure the questions were comprehensible. Teachers administered it during school hours for a maximum of forty-five minutes.

Table 2 Sample questions, sub-questions, and answers from the online questionnaire
Full size table

The sequential application of data collection methods enabled the research team to develop guides for the focus groups and individual interviews based on quantitative data and initial analysis. As a result, three interview guides were elaborated, considering the age of participants, sub-samples, and role in bullying. Before data collection, these interview guides were tested on same-aged volunteers to ensure they were easy to understand and provided accurate answers. These guides were similar in their open-ended questions and, in line with our research question, they focused on: (a) the young people’s own experiences of bullying (these experiences could be linked to different roles of bullying in order to explore the diversity of bystanders’ experiences and account for the permeability of roles that is assumed in this study), (b) their behaviors, attitudes, and cognitions and those of other peers (to explore social influence), (c) their relationships with peers and how these evolved throughout bullying (cf. Table 3). The guides were designed so young people were asked certain sub-questions based on their previous answers to probe them further. Before the focus groups and individual interviews, the participants’ personal information (age, gender, number of siblings, etc.) was collected using fill-in-the-blank cards or talking to one or more adults in the school.

Table 3 Sample questions and sub-questions from the interview guides (for focus groups and individual interviews)
Full size table

Individual interviews and focus groups were conducted by two research team members (one facilitating and one observing) or by one research team member and one psychologist if the latter participated in the discussion (in this case, the researcher facilitated the exchanges). A maximum of six teenagers could participate in a focus group. These focus groups were conducted in rooms made available by the school during school hours. The individual interviews were conducted outside school hours in rooms provided by psychologists or in locations chosen by the respondents. All the exchanges lasted one hour and thirty minutes and were audio-recorded.

Ethical Considerations

The research project related to this study was subject to ethical review in 2018 by a commission appointed by the University of Teacher Education Valais to ensure compliance with the research in schools and ethical guidelines on children’s research (Moody & Darbellay, 2019; Moody et al., 2021; Morrow, 2008). The objectives of the project were presented to participants in age-appropriate language by an adult known to them (teacher, school mediator, psychologist, or adult of trust) and in writing. Participants received an information letter about the research project, which they passed on to their legal guardians. All participants who wished to participate in the project and those from classes selected to complete the online questionnaires received this letter. A consent form was enclosed in the letter. This form had to be fully completed and signed by the participants and at least one legal guardian to allow participation.

Anonymity was ensured using pseudonyms chosen by the participants themselves. Additionally, the confidentiality of participants’ responses was guaranteed. A code was also randomly assigned to the class of the participants. Participants could thus withdraw from the project at any time, without prejudice or negative consequences, and retrieve their data. Moreover, all participants were free to consult their data whenever they wished. These data were deleted at the end of the project. The research team also ensured not to use data that could identify or isolate a particular school class or young person. To reduce secondary victimization, participants were offered practical advice to discuss their problems (information leaflet, adult counselor, etc.). Ultimately, findings were communicated to them in an age-appropriate manner.

Data Analysis

Although quantitative data are used to support qualitative ones in this study, they were first collected to find out quickly, from the responses of many participants, if there is social influence among them and to observe their implications in terms of relative frequencies on their behaviors, attitudes, and cognitions, as well as the reasons for their emergence in relation to peer relationships. Quantitative data analysis was then used to guide an initial qualitative data analysis. Then, in line with a comprehensive and iterative approach (Charmillot & Seferdjeli, 2002; Moseholm & Fetters, 2017), the qualitative data analysis led the research team to delve deeper into the quantitative data analysis. Thus, the analysis of quantitative and qualitative data was carried out in a systemic logic, involving continuous back-and-forth, constant integration and synthesis between data collected, analyzed, and even re-analyzed as new information on bullying was gathered from participants as suggested by Hong and Espelage (2012). Answering the research question prioritized qualitative data, enabling researchers to understand better and describe the implication of social influence and peer relationships on bullying bystanders based on the meaning they attribute to it.

Quantitative data were analyzed using IBM SPSS Statistics 23.0 software. A descriptive analysis was performed using the ‘complex sample’ module. This analysis allowed, among other things, the observation of whether certain behaviors, attitudes, or cognitions (i.e., feeling obliged to do or say something bad to peers, the importance of thinking the same as a peer group, etc.) and the reasons for their emergence (i.e., to be like a schoolmate or schoolmates, to be accepted by schoolmates, etc.), were displayed more frequently than others. Each behavior, attitude, cognition, or reason was treated as a separate variable to estimate a frequency relative to its value.

The qualitative data analysis was performed in several phases using MAXQDA software (2018). These phases are derived from three methods regularly used to analyze qualitative data, consolidating the applied approach in this research: the grounded theory of Strauss and Corbin, the qualitative analysis method of Miles and Huberman, and the thematic analysis of Paillé and MucchielliFootnote 2. The purpose of these phases is to establish categories (or themes), which are then compared to each other to provide a deeper understanding of links between items (Intissar & Rabeb, 2015), here between social influence and bystander’ relationships with other peers.

The first phase of the qualitative data analysis was transcribing the recordings of the individual interviews and focus groups (verbatim). The second phase involved using the transcriptions to break down the participants’ utterances into units of meaning. Each unit of meaning was then associated with one or more keywords (i.e., popular young people, close friend, does not have friends, fear that peers will turn against them, fear of siding with victims, is part of a peer group, is not part of a peer group and trying to join). In the third phase, keywords with common properties were grouped to create categories (i.e., obedience, conformism, compliance, and social influence at an interpersonal or intragroup level). The last phase involved linking and comparing these categories using strategies such as adversarial resolution (identifying opposing categories and emerging new knowledge based on the resolution of the opposition) or classification (classifying categories according to their recurrence; Intissar & Rabeb, 2015).

Results

Given that this study is based primarily on qualitative data, with support from quantitative ones, the results are presented according to a qualitative logic. They are thus structured in categories derived from the data analysis and linked to theoretical knowledge on social influence and peer relations, in line with the challenges of a comprehensive and iterative approach described above.

Different Forms of Social Influence

Conformity

Quantitative and qualitative data show that conformity is the most common form of social influence among young people confronted with bullying. It is mainly exercised on teenagers’ attitudes and cognitions, including those who are bystanders of bullying. For example, Julie (girl, 12–14 years old) said during a focus group: “I don’t know her, but I don’t like her because lots of people have said bad things to me about her”. Similarly, Lanny (boy, 12–14 years old, focus group) explained to a friend he used to have: “Yeah, well, I agree with you, like we used to be mates, but then there were so many rumors about you all the time, every day, it’s hard not to believe them”. Quantitative data also emphasize that sharing specific cognitions with other peers is essential for students (cf. Table 4, variable 1, question (hereafter: Q): It is important to you that your schoolmates think the same as you?).

Table 4 Relative frequencies
Full size table

While qualitative and quantitative data reveal the impact of conformity on teenagers’ cognitions and attitudes, quantitative data suggest a lack of conformity among peers regarding their behaviors or that they are less aware of it. Indeed, very few admit feeling pressure to do or say mean things to a schoolmate (cf. Table 4, variable 2, Q: Have you ever felt pressured to do or say nasty things to a schoolmate? ). Nevertheless, when cross-referencing quantitative with qualitative findings, oppositions emerge. Resolving these oppositions suggests the need for higher conformity to change bystander behaviors, for example, by encouraging many other young people to adopt similar behaviors: “I think that I followed the others because, well, everyone was bullying her, so I did too. I didn’t really know her. I hardly knew who she was, well, to be fair, I didn’t know her at all” (Baptiste, boy, 12–14 years old, focus group); “I think they just have to back up their mates, one of them says something, and they all laugh” (Juliette, girl, 12–14 years old, focus group). More specifically, these words show that conformity can thus take a more indirect form by creating social norms, which may explain why some bystanders are unaware of the impact of social influence on their behavior. The more behaviors become social norms (through their increasingly frequent use), the more likely bystanders will adopt them.

When bystander students, who are aware of the impact of conformity pressure on their behavior, are asked why they have felt pressured to do or say nasty things to another peer [schoolmate], most of them say that they do not know why (cf. Table 4, variable 4, sub-question (hereafter: SQ): Why have you ever felt pressured to do or say nasty things to a schoolmate?). This confirms questions about awareness of the impact of conformity on teenager’s behaviors; even when they are aware of it, they don’t necessarily manage to explain it. However, as shown in Table 4 (cf. variable 4), other reasons explain conformism. The most common is “fearing that peers [schoolmates] would be angry with them” and “to be like a peer or peers [schoolmate or schoolmates]”, confirming the need for bystanders to be socially accepted and recognized by their peers through social rewards. This also highlights that social influence can be higher when bystanders fear being bullied by their peers. More broadly, these results indicate that conformity can be linked to, and may even be exacerbated by, other forms of social influence, such as compliance.

Compliance and Obedience

By giving voice to young people, qualitative methods and related findings provide a more fluid perspective on bullying, especially on adolescents’ behaviors, attitudes, and cognitions in this respect. Qualitative data have deepened our understanding of the phenomena of compliance and obedience. It should be noted that, in this study, quantitative data do not support a detailed understanding of compliance and obedience, except when they are discussed with student popularity, as shown below in the analysis of variable 5 of the survey (cf. section on Social Influence at an Intergroup Level).

Qualitative data highlight the fluidity within bullying and confirm that compliance can generate a role permeability among young people. Indeed, it is relatively common that they switch from one role (bystander, victim, perpetrator) to another (Moody et al., 2020). Nono (boy, 12–14 years old, focus group) explained that he was socially rewarded (e.g., gained acceptance) by his peers when he bullied another victim: “I got out of it by doing something really mean, well, they were hassling me […] and to stop them bullying me, I had to become a bully and then I was on their side”. From a more psycho-sociological perspective, this example indicates that the boundaries between victim and perpetrator roles are permeable, and that of the bystander is transient and, therefore, difficult to maintain in the medium to long term. Therefore, bystanders may have to conform to perpetrators and thus seek social rewards, or at least not to lose any (compliance), to protect themselves. Albane (girl, 12–14 years old, focus group) recounts that she lost social acceptance from her peers when she sided with a victim: “There was one girl who was being bullied. I stood up for her, and then I was insulted because I’d done that”. This highlights the complexity and risks involved in the bystander role, whose sustainability needs to be questioned.

Concerning obedience, qualitative results, which highlight the fluidity in adolescents’ behaviors, attitudes, and cognitions, do not separately support the consideration of this form of social influence. Instead, results indicate that obedience emerges only in interaction with other forms of social influence. Several young people confirm that they might become obedient because they fear losing social rewards (perceived popularity), which links obedience directly to compliance. Mo-Avan (girl, 12–14 years old, focus group) confirms: “I didn’t want to turn her against me because she was the most popular girl in the class, and I liked her anyway. If she turned against me, she could turn all my other friends against me”. Similarly, as a means of protecting himself, Lu (boy, 12–14 years old, focus group) admits that it is preferable as a bystander to side with popular students, which implies that obedience could facilitate role changes in bullying: “Bystanders join up with the popular student, they always go with the strongest, you know”. Two points should be highlighted: first, as explained below, the interplay of obedience with other forms of social influence vary according to peer relationships in which they take place. Second, according to Mo-Avan, perceived popularity does not exclude a more sociometric one.

Interactions with Different Forms of Relationships

Social Influence at an Intergroup Level

Qualitative and quantitative data suggest that the most common level of social influence is exerted between peers belonging to different groups: intergroup influence. Data thus confirm our first expectation, adding another level to those identified in the literature on bullying (intragroup or interpersonal). According to Luana (girl, 12–13 years old, individual interview), bystanders conform their cognitions, attitudes, and behaviors to those of the group members they wish to belong to, hoping to meet their expectations. She describes how conformity takes place: “Students act like the others so they can join their group. For example, if they want to join the popular kids, they will act just like them; if they insult someone or favor pollution, they will also favor pollution. They won’t have their own opinions”. Luana’s words are supported by quantitative data, which indicate that when bystander students felt compelled to hurt a peer, it was primarily because they were socially influenced by peers outside of the group they belonged (cf. Table 4, variable 3, SQ: Who are schoolmates for whom you have felt pressured to hurt another peer?).

As Luana says, the reason why bystanders conform to the perceived expectations of members of the peer group to which they wish to belong is closely linked to compliance as a form of social influence. By complying with expectations, youth bystanders increase their chances of being accepted and recognized by a group. At school, students feel a social constraint to belong to at least one group: “It’s just that you have to force yourself to be in a group, it’s a bit awkward if you don’t have any friends” (Sarah, girl, 12–14 years old, focus group). Such pressure also relates to the basic needs of safety and survival in a social and sometimes hostile environment; as Francis (boy, 12–14 years old, focus group) puts it: “I feel like, in 2019, if you don’t fit in, you’re dead […]”. Additionally, if all students belong to a group, the social norm will exert an increasing influence on them, leading to a dynamic of continuous reinforcement and the strengthening of social cohesion among them.

Qualitative data suggest that compliance can be reinforced by obedience to dominant peers, particularly at an intergroup level. This finding further supports previous observations (cf. section on Compliance and Obedience) by confirming that obedience at this level is strongly linked with perceived popularity and may intersect with compliance. Additionally, it implies that to achieve a sense of security and to survive in a social context, it is even preferable for bystanders to seek to belong to groups whose members enjoy perceived popularity. As participants admit, seeking to create this kind of relationship can have a protective effect: “The others are afraid of them [popular people], so they’ll try to be friends with them” (Lotte, girl, 12–14 years old, focus group); Leading bystanders to try to gain social rewards from popular members, out of fear.; “[… with] popular people, […] you’re half clueless let’s put it that way and half popular’ ” (Luana, girl, 12–13 years old, individual interview).

Social Influence at an Intragroup Level

While social influence mainly comes from outside the peer group, qualitative and quantitative data reveal that young people also influence each other when they have relationships. In this respect, conformity can occur within peer groups, i.e., at an intragroup level. Qualitative findings underline that this in-group conformity can lead to the emergence of both prosocial and violent attitudes, cognitions, and behaviors in bystanders.

For example, Noël (boy, 17 years old, individual interview) says how members of the same group conformed as bystanders to support him when he was bullied: “After one of them came over to talk to me and help me, suddenly, all his mates followed him, a bit like sheep, also to listen and help me”. Baptiste (boy, 12–14 years old, focus group) also confirms: “We were in a group, so when someone said ‘Yeah, there’s no point [to pick on her or him], it’s mean’. At once, all the others, me included, supported this idea”. Conversely, Kassandra (girl, 12–14 years old, focus group) describes how bystanders may conform to other members of their group by participating in, reinforcing, or imitating bullying: “I was as afraid that she [the one involved in the violence] was going to talk with her friends and that other people I didn’t know might turn against me [if I didn’t do as they did]”. These accounts reveal that the group norms can swiftly change when a member – even a bystander – agrees or disagrees with bullying. This change can lead to a rapid deterioration or, conversely, improvement in bullying or even to its cessation. However, to promote conformity and primarily influence bystander behaviors in a group context, the members inciting change must have a high degree of influence over their peers, which also depends on the group’s social cohesion.

Social norms can evolve even quicker within a group if conformity between members is combined with compliance and/or obedience. Interestingly, intragroup obedience is not expressed the same way as at an intergroup level, suggesting that it no longer relates to compliance (although it remains linked to conformity). Inside a group, popularity does not necessarily come from the fear generated among members, which tends towards a more sociometric form of popularity. In the questionnaire, the 152 bystanders who reported experiencing pressure to harm someone from a group member were asked to identify who exerted this pressure. A large proportion answered that such pressure came from peers who were “listened to” within the group, translating to high levels of sociometric popularity (cf. Table 4, variable 5; SQ: Which peers in your group(s) forced you to harm?). Others also responded that the influential members were the ones who tried to maintain peace in the group or who advocated for the ideas of the most listened to. Therefore, popularity in a group can also be closely supported by the capacity to maintain social functioning among members, giving rise to a relatively new form of popularity known as “functional popularity”. This is confirmed by additional responses, which indicate that members with the most minor social influence hinder the group’s functioning or cannot maintain it. According to questionnaire respondents, these members disagree with what the majority thinks or does, do not speak much, prefer being alone, and are bullied or teased by peers. Thus, quantitative findings show that adolescents most at risk of bullying in a group are those who do not respect the functioning and social norms in place.

Social Influence at an Interindividual Level

Giving voice to young people makes it possible to observe their use of distinct words to describe different intensities of relationships (close/privileged relationships: friend; less close: buddy, regular friend, etc.). This is more challenging to discern through quantitative results, which is why the latter does not contribute to explaining social influence at the interindividual level in this study.

Qualitative data show that privileged relationships with victims can encourage bystander young people to help them: “Loads of people say ‘Yeah, look at that ginger one’, and pick at my friend. Then, because he’s my friend, I talk back to them” (Nono, boy, 12–14 years old, focus group). More to the point, what motivates bystanders to help victims is the intensity of their privileged relationships, such as being best friends. Xy07 (girl, 12–14 years old, focus group) confirms: “[…] My best friend was being bullied by text […]. Once, she got this voicemail telling her she had to kill herself […], so she’s my best friend, and I was really scared that she was going to do something, so I got involved and told the teacher […]”. This relational intensity, as intrinsic motivation, affects bystanders enough to counterbalance conformity, even in social contexts where violence is valued.

As explained above, the role of bystanders carries several risks. However, bystander relationships with peers can offset these risks, particularly if they are friends. This is confirmed by Nono (boy, aged 12–14, focus group), who said that if bystanders who assist victims are his friends, it is less likely to be a problem for him: “ Well, see, if I see Paul, my mate, it’s no problem if he hangs out with people like that [talking about a seriously bullied girl, who he personally despises and perceived as provocative in school]. But if I see someone I don’t really know with her, I’ll be disgusted!”. This confirms that when bystanders express behaviors, cognitions, or attitudes that deviate from their expectations (compliance), their friends are likelier to tolerate and accept these differences without losing their social rewards. Consequently, bystanders are more inclined to engage in prosocial behaviors, attitudes, and cognitions toward victims when they have multiple privileged relationships, particularly those with high intensity.

However, privileged relationships can also contribute to the emergence of violent behaviors, attitudes, and cognitions in bystanders. When relationships between bystanders and victims are not sufficiently intense, conformity or compliance may prevail among bystanders but be directed toward peers rather than victims. On the one hand, Rose (girl, 13 years old, individual interview) explains how a girl she was friends with was influenced to conform as a bystander to other social norms, including accentuating violence towards her: “She started blanking me when she had her new best friend, she started insulting me and all that”. On the other hand, Lotte (girl, 12–14 years old, focus group) recounts how two teenagers pretended to be friends with her to get information and reinforce violence against her to meet the expectations of other peers: “They [peers] kind of took me under their wing just to get information from me, and now this year they’re talking crap about me […].” This highlights that compliance is not always reciprocal among young people. This non-reciprocity also raises questions about obedience within privileged relationships. Is it simply that qualitative data in this study do not reveal this social influence within these relationships? Or does it take a different form in relation to inter- and intra-group levels, possibly linked to more subtle dynamics of domination closely associated with relational intensity?

Discussion

The main research question of this study is to explore how different forms of social influence and peer relationships intersect and affect student bystanders of bullying. Qualitative results of this study, supported by quantitative ones, have led to several conclusions, which are discussed in the following three parts. Each of these parts examines a specific aspect of the research question in detail.

Interactions between Different Forms of Social Influence

Qualitative and quantitative results of this study highlight that conformity is the most prevalent form of social influence among bystanders, with varying effects depending on whether it manifests through direct observations or social norms. These qualitative and quantitative results validate earlier studies that highlight the impact of social norms on conformity (Harkins et al., 2017; Laninga-Wijnen et al., 2020; Stahel & Moody, 2023; Veenstra & Lodder, 2022). However, current qualitative findings go beyond previous studies by shedding light on the pivotal role of social norms in shaping bystanders’ behaviors. While direct observations of bullying can shape bystanders’ cognitions and attitudes, actual behavior changes, such as reinforcing or imitating bullying, require a more nuanced approach to social norms. Thus, for bystanders to reinforce or imitate bullying, many other peers must engage in similar behaviors, or various forms of social influence (compliance or obedience) must combine with conformity to encourage such behaviors. As behavior change requires additional effort, this highlights the importance, as Howe and Krosnick (2019) point out, of further investigating whether attitudes and cognitions should support behavior change.

In addition, the mainly qualitative results of this study underscore the interplay between different forms of social influence, some of which appear only in conjunction with others, such as obedience. On the one hand, examination of these interactions provides a better understanding of Olweus’s (1991) conception of the fixed, enduring roles of bullies while at the same time highlighting the permeability of these roles, with the role of bystander being mainly transitory. Associated with conformity and/or compliance, obedience can facilitate role changes when bystanders struggle to meet their peers’ expectations. Data point in two main directions to explain why obedience appears systematically in conjunction with other forms of social influence. On the one hand, qualitative results reveal that bystanders adopt the role of perpetrator to assert themselves against their peers when conformism and compliance are insufficient, as a last-resort strategy to protect themselves from becoming the next victim of bullying.

On the other hand, before this ultimate solution is reached, qualitative results lead us to believe that the transition to the perpetrator role occurs gradually among bystanders because of perceived popularity. This deepens our understanding of popularity and its impact on bystanders, as studied by Brown and Larson (2009), Kindelberger (2018), and Salmivalli (2010). Due to failing their peer expectations due to conformity or compliance, bystanders will gradually endorse the attitudes, cognitions, and behaviors of peers perceived as popular (obedience). Under social peer pressure, these popular youth become guardians of the social norms to be respected, and by obeying, bystanders increase their chances of acquiring dominance or salience, transitioning towards the role of perpetrator.

Interactions between Different Forms of Social Influence and Relationships

The quantitative and qualitative findings demonstrate that while conformity is more prevalent among young people during bullying, it is predominantly observed at an intergroup level when bystanders seek to create new group relationships with peers. However, this does not exclude the presence of conformity in other existing relationships (group or privileged relationships). Firstly, these quantitative and qualitative results emphasize the importance of incorporating the intergroup level into the study of social influence and its impact on bullying bystanders, in addition to the intragroup and interpersonal levels already identified in research (Brown & Larson, 2009).

Secondly, the qualitative findings clarify the observation made by Laninga-Wijnen et al. (2020) that teenagers show greater tolerance towards peers who deviate from social expectations or norms when they share an intense relationship. Nevertheless, since conformity is most pronounced outside existing relationships, it is interesting to underline that even less intense relationships (group or privileged) can be characterized by some form of tolerance. This explains why popularity can take on a more sociometric form within existing groups and suggests that quality identity development can occur in these relationships and less intense privileged relationships, as freedom of exploration would already be possible (Ragelienė, 2016). This does not rule out the possibility that tolerance between young people can increase as their relationships intensify when they become friends. Ultimately, it appears that when bystanders do not reinforce or imitate bullying, in the case of prevailing social norms valuing such violence, it is primarily the relationships they aim to create with other peers, especially group ones, that are affected.

The qualitative and quantitative results of this study indicate that conformity and obedience are mainly observed in relationships involving several young people (intergroup and intragroup levels), confirming that these forms of social influence are motivated by the need for recognition and social acceptance. These results complement the Influence-Compatibility model (Giletta et al., 2021; Laninga-Wijnen et al., 2020; Laursen & Veenstra, 2021; Veenstra & Huitsing, 2021) by clarifying that bystanders’ search for similarity goes beyond the simple formation of a relationship, since it serves to ensure their survival in the social context. Noteworthy, the need for survival is more pronounced at the intergroup level, where bystanders struggle to create relationships and face more significant risks of non-compliance and lack of conformity, than at the intragroup level, where relationships and similarities are already established between peers. Moreover, as explained above, in intergroup contexts, when conformity and compliance are insufficient to gain recognition and social acceptance from other peers, obedience may manifest itself in harsher forms, such as dominance and the pursuit of perceived popularity among bystanders. On the other hand, in intragroup contexts where relationships are already created, popularity can take on a more sociometric form, as the survival of bystanders is less at stake.

The quantitative results also reveal a complementary form of mainly functional popularity that emerges within intragroup dynamics. This form of popularity complements the two forms previously identified in the literature (sociometric or perceived popularity; Brown & Larson, 2009; Kindelberger, 2018; Salmivalli, 2010). Functional popularity characterizes bystanders and other young people who actively seek recognition and acceptance from their group members by promoting social cohesion within the group and, by so doing, ensuring their own and other peers’ survival in the social context.

Regulating Social Influence through Friendships

The qualitative results of this study provide two additional insights into the study of the links between social influence and privileged relationships. Firstly, qualitative results highlight the dynamics between bystander students and victims when their privileged relationships are not sufficiently intense or non-existent. In this case, bystanders may reinforce or imitate violence towards victims, either to conform to existing social norms (if they privilege violence, conformity) or to gain social rewards from other peers (conformity). These qualitative results challenge the notion of reciprocal social influence proposed by Brown and Larson (2009) and highlight a more unilateral form of social influence. Consequently, it raises the question of whether incorporating other types of relationships in studying social influence would allow us to observe more unilateral or asymmetrical forms of social influence.

Complementing the first insight, the second, based on qualitative findings, emphasizes that friendships can regulate various forms of social influence and encourage mutual help between young people. The intensity of relationships is thus a driving force that outweighs the need for external social influence to change the attitudes, cognitions, and, above all, behaviors of bystanders. These qualitative findings further elaborate on Laninga-Wijnen et al. (2020) suggestion that prosocial friendships are less influential than privileged relationships that prioritize violence. This can be observed in bystanders whose privileged relationships with victims are inexistent or insufficient or in those who aim to create relationships with peers who privilege violence. However, in the case of friendship between bystanders and victims, the intensity of their relationship can mitigate the impact of violent social norms (conformity) and encourage the adaptation of more prosocial behaviors, attitudes, or cognitions towards bullied peers. Conformity may also be regulated if bystanders have multiple friendships with peers and, thus, do not fear as much being judged by some if they side with the victims.

Finally, it should be noted that the results of this study (both qualitative and quantitative) do not reveal whether privileged relationships with perpetrators have similar effects on bystanders as those with victims, i.e., taking their side or turning against them depending on relational intentionality. Further research is needed to investigate this aspect and obtain a complete understanding of relational intensity, whether in privileged or other forms of relationships, and their influence on bystanders.

Awareness, Prevention, and Intervention Implications

The results show that it is essential to consider different forms of social influence in designing and implementing anti-bullying actions. As these social influences manifest differently depending on the relationships young people already have or aspire to develop, professionals should plan diverse actions at various levels (prevention, awareness, intervention, etc.). These actions should be guided by the following insights derived from the study results, including:

As social influence, notably through conformity, emerges mainly between peers who do not have existing relationships but aspire to form them (intergroup level), findings suggest that in a youth environment, it is central to work with most of them upstream of bullying situations. For example, promoting the development of socio-emotional skills among young people, especially towards victims (see Hoareau et al., 2017; Stassin & Lechenaut, 2021). A high level of conformity may emerge from interactions between the adolescents involved, increasingly encouraging them to adopt prosocial behaviors, attitudes, and cognitions. This prevention could positively affect all levels of social influence (intragroup, intergroup, and interindividual) and social norms among teenagers. Moreover, the emergence of prosocial attitudes, cognitions, and behaviors would be valued if teenagers socially rewarded their peers for their prosocial actions (compliance) and even more if they obtained more significant sociometric popularity at an intergroup level or functional popularity at an intragroup level (obedience).

Several actions could complement this prevention. Because social influence is closely related to existing social norms among teenagers, which may vary from one environment to another or even from one group of peers to another (Laursen & Veenstra, 2021; Stahel & Moody, 2023; Veenstra & Lodder, 2022), it would be interesting to analyze norms in place in a particular environment (such as in a classroom) or group to provide those who value violence with prevention actions that are more focused on the emergence of prosocial attitudes, cognitions, and behaviors.

In addition, based on the permeability results, when an environment is composed of multiple young people and is at risk for the emergence of bullying, these youth could benefit from the explanation of bystander roles, both its benefits and related risks. These risks could emphasize the importance of young people talking about bullying to avoid becoming either a perpetrator or a victim in the medium term. At the same time, to promote a more sociometric form of popularity and through this self-help, the effect of peer feedback and social rewards could be explained to young people in relation to their identity development to encourage positive social evaluation among themselves, specifically when they adopt prosocial attitudes, cognitions, and behaviors (Laursen & Veenstra, 2021; Somerville, 2013; Van Hoorn et al., 2016). Through such feedback, young people would define their identity in prosocial ways of doing, being, or thinking. This could also help create more privileged or even more intense relationships between them and increase their chances of being helped in the event of bullying.

Although many current methods (e.g., Support groups and Shared concern methods; see Senden & Galand, 2021) advocate working with a group of youth bystanders to intervene in and stop bullying situations, it should be straightforward that depending on the degree of support or prevention that teenagers receive, they will make mistakes. More specifically, the findings show that certain aspects must be respected when composing this group to reduce these mistakes. Firstly, this group should be composed of many bystanders so that they can conform to each other in more prosocial attitudes, cognitions, and behaviors toward victims, specifically if these teenagers are not accustomed to these ways and/or if these ways are not valued among themselves. Secondly, to support the orientation of conformity toward prosocial behaviors, attitudes, and cognitions toward victims, it is helpful to include bystanders who have intense relationships with victims (or at least not with perpetrators) in the group. Thirdly, conformity will be more likely to be directed toward prosocial attitudes, cognitions, or behaviors if bystanders socially reward each other for adopting them (compliance) or if, through these ways, they develop sociometric and/or functional popularity (obedience). Perceived popularity must be avoided to prevent a repeat of the violence. If any of these three aspects cannot be met, implementing only one is certainly possible, but with careful guidance. This would enable bystanders to reduce their risk of reprisals from other peers and thus limit the shift to the role of victim or even perpetrator.

Limitations and Future Studies

Several limitations can be noted. All are closely related to the need to understand better interactions between different aspects of bullying (social influence, relationships, etc.), particularly those concerning bystanders. This suggests the need for an even more fluid understanding of bullying than that proposed in this study.

First, studying the links between social influence and relationships between bystanders and their peers could gain explanatory power if their interactions were examined in depth, either by broadening the range of forms studied or by directly soliciting participants’ views on the types of social influence and relationships they observe. To fill the gaps identified in the discussion section, it would be interesting to investigate whether social influence between young people is always reciprocal or can be unilateral (as the results of this study suggest), and studying its links with peer relationships would improve our understanding of the dynamics of groups membership (Brown & Larson, 2009). In addition, a more in-depth study of privileged relationships would provide a better understanding of their association with similarity seeking and their impact on bystanders when formed with the perpetrator. Finally, more data on the effects of relational intensity on social influence relationships other than privileged ones would be insightful.

Secondly, this study would benefit from investigating the interplay between social influence and relationships among bystanders and their peers while considering other aspects. Aspects such as age, grade level, and duration of relationships would provide valuable insights to inform anti-bullying actions. For example, this would clarify the above discussion on the composition of bystander groups participating in interventions.

Finally, although this study shows the effects of social influence and relationships on bystanders’ prosocial behaviors, attitudes, or cognitions, it should be combined with others on the development of prosocial ways, especially in adolescence and when relationships are not necessarily established between peers, or they do not know each other (Eisenberg et al., 2010; Van Rijsewijk et al., 2016). These findings would further support the practical implications proposed above.

Conclusion

This study aims to deepen our understanding of bullying by examining different forms of social influence through the relationships of bystander students during adolescence.

The results of this study highlight several strong points. First, conformity is the most common form of social influence among bystanders, particularly when they aim to create group relationships with other peers. However, it is noteworthy that conformity can intersect with other forms of social influence, such as compliance and obedience. This is particularly true when conformity alone fails to ensure that bystanders perceive themselves as similar to their peers and does not sufficiently contribute to creating and sustaining their relationships. Secondly, results on obedience confirm that popularity can manifest itself in various ways depending on the relationships between young people; a relatively new form of popularity (functional popularity) emerges within peer groups. Finally, privileged relationships with victims, particularly those marked by strong intensity, can regulate social influences on bystanders and motivate them to provide support. This highlights the importance of professionals considering the relationships between peers and their intensity, in interplay with social influence, when designing and implementing concrete actions to address bullying.