Bullying is understood to be a social problem that has severe negative impacts extending beyond the outcomes for those who are directly involved, also affecting bystanders, families, and communities (Salmivalli 1999). Research has consistently shown that bullying is a destructive phenomenon that is particularly prevalent during late childhood and early adolescence which can result in long-term detrimental, and sometimes fatal, outcomes (Hawker and Boulton 2000; Stassen Berger 2007; Wolke et al. 2013). It is a behaviour that is found to be expressed by young people in virtually all schools regardless of contextual and geographical differences and can be seen as one of the most pressing modern public health concerns for children and adolescents (Espelage and Swearer 2003).

There are different types of bullying, which are commonly classified as either direct (overt) or indirect (covert) (Van der Wal et al. 2003). Forms of direct bullying include physical (kicking, pushing, and hitting) or verbal acts (repeated derogatory remarks and name-calling) used to hurt or humiliate the victim (Olweus 1993). In contrast, indirect forms of bullying are not always carried out in front of the victim, and often include or can occur via a third party, such as spreading rumours and social exclusion (Rivers and Smith 1994; Stassen Berger 2007). With the rise of technology and social media over the past decade, a newer form of bullying has also emerged: cyberbullying—which uses electronic means (for example through texting, messages, or posts on social media) to inflict harm on others (Slonje et al. 2013).

Although we know a lot about different manifestations of bullying, both research and interventions targeting bullying tend to be fraught with definitional issues and inconsistencies in measurement. Such problems with defining the core construct of bullying and how it is experienced by young people have created inefficiencies in our ability to generalise independent research findings. The following study, therefore, seeks to examine the issue of defining bullying in greater detail, specifically investigating how adolescents understand and experience bullying as compared to scholarly definitions of the same behaviour.

Problems in Defining Bullying

Bullying has been the subject of scientific study since the term ‘mobbing’ was first introduced in the context of racial discrimination during the 1970s (Olweus 1979, 1995). Olweus (1993) later developed one of the most commonly used definitions of this behaviour as applied to the schooling environment. Specifically, he suggested that bullying existed ‘when he or she (a student) is exposed, repeatedly and over time, to negative actions, on the part of one or more other students’ (Olweus 1993, p. 9). Olweus further defined such ‘negative actions’ as including intentional acts such as inflicting injury or discomfort, through either verbal or physical means. However, over the past 30 years of research, the definition of bullying has been inconsistently applied, and much debate has occurred regarding the distinct manifestations and measurement of these behaviours (Huang and Cornell 2015; Vaillancourt et al. 2008). It is only recently that the Center for Disease Control and Prevention (CDC) and the US Department of Education (ED) released a comprehensive definition of bullying in the hopes this would increase consistency:

bullying is any unwanted aggressive behaviour(s) by another youth or group of youths who are not siblings or current dating partners that involves an observed or perceived power imbalance and is repeated multiple times or is highly likely to be repeated. Bullying may inflict harm or distress on the targeted youth including physical, psychological, social or educational harm (Gladden et al. 2014, p. 7).

Three main features (repetition, harm, and an imbalance of power between the bully and the victim) as defined by the original framework of Olweus (1993) are still incorporated in this definition as key characteristics of bullying. Yet notably, the intentionality of negative acts is missing from the CDC definition and the elements of harm and imbalance of power have been expanded. Specifically, harm is defined as multidimensional (physical, social, emotional, or educational), and power imbalances are also no longer dependent on just physical strength, but may include both objective and subjective elements of social or psychological strength (Raskauskas et al. 2010). Additionally, aggressive behaviours are defined as unwanted (but not necessarily intentional), and while repetition of the behaviour is a core factor, the definition allows for a single act of aggression to be labelled bullying if it is perceived to have a high likelihood of reoccurring (Vivolo-Kantor et al. 2014). These issues illustrate a fundamental tension in examining bullying whereby research, policy, and legislation attempt to report and measure these behaviours in objective ways, but due to the subjectivity of an individual’s experience, this may not reflect the experience of youth themselves. The CDC definition offers some scope to expand our understanding and measurement of bullying by suggesting that bullying involves a mediation such that the victim must consider the behaviour to be ‘unwanted’ in order for it to be bullying.

Despite these positive changes and the growing consensus on the underlying characteristics of bullying, issues still remain with respect to the understanding and application of researcher-developed definitions in the broader population (Vivolo-Kantor et al. 2014). In fact, a number of research studies have found that students ascribe different meanings to bullying than those in scholarly definitions as the aforementioned (Arora 1996; Hellström et al. 2015). Specifically, both repetition (Land 2003; Naylor et al. 2006) and power imbalance (Vaillancourt et al. 2008) are often found to be absent from students’ accounts of bullying. Furthermore, studies have shown that students have a tendency to constrain their definitions of bullying to direct forms of physical or verbal bullying (hitting, punching, name-calling, threatening, and coercion), rather than indirect, relational forms of bullying such as exclusion (Boulton et al. 2002; Naylor et al. 2006).

These differences in conceptualisations of bullying may be a result of developmental changes in young people’s definitions of this behaviour over time. For example, Smith et al. (2002) found that whilst older children and adolescents (11 and 14 years old) were able to distinguish between distinct forms of aggression, physical violence, and bullying, children aged 8 years old still considered fighting (between equally matched peers) as well as an array of other negative interpersonal behaviours to be bullying. The older children were also able to articulate the concept of relational bullying (e.g. exclusion), and the adolescents included both indirect bullying in their accounts and imbalances of power. Hellström et al. (2015) replicated these findings, showing that younger children are often unable to distinguish between broad types of aggressive behaviours and bullying. This suggests that the understanding of bullying may develop over childhood, although this does not fully account for the discrepancies between young people’s understanding and experiences of bullying as compared to academic conceptualisations.

Recent reviews have suggested that adolescence is the peak age period in which individuals engage in not only traditional forms of bullying but also in cyberbullying (Slonje et al. 2013; Tokunaga 2010). According to Tokunaga (2010), 20 to 40% of young people have experienced cyberbullying at least once. The prevalence and significance of cyberbullying as a distinct form of bullying have been demonstrated consistently (Grigg 2010; Slonje et al. 2013) and should be considered as equally, if not more, pervasive and detrimental as traditional forms of bullying (Mishna et al. 2009). The phenomenon of cyberbullying has also had no shortage of academic attention and debate regarding its conceptualisation and definition. Researchers have questioned the relevance of Olweus’ criteria within the virtual context and have proposed the addition of two criteria (anonymity and public versus private) with mixed success (Langos 2012; Slonje and Smith 2008; Vandebosch and Van Cleemput 2008).

One study comparing cyberbullying definitions among young people across six European countries found that of the five criteria (intentionality, repetition, imbalance of power, anonymity, and public versus private), imbalance of power, intentionality, and anonymity, respectively, were considered most central when defining cyberbullying (Menesini et al. 2012). That is, participants predominantly considered traditional bullying criteria, with the exception of repetition, when evaluating behaviours as cyberbullying. Unlike the public versus private criterion, which did not show any relevance for the definition of cyberbullying, anonymity was found to influence cyberbullying definitions when considered in conjunction with intentionality. Results showed a higher probability of defining actions as cyberbullying if they were intentional and non-anonymous and a lower probability if the act was non-intentional and anonymous. Conversely, another study investigating cyberbullying from the perspectives of young people found that participants reported cyberbullying as the worst and most serious form of bullying because of the perceived anonymity of the perpetrator. Anonymity was seen to increase the distress of the victim and acted as a barrier against seeking help due to the perceived lack of evidence of how to identify bully (Mishna et al. 2009). Whilst some researchers have gone a step further and called for an alternative, broader approach that focuses on the wide array of behaviours known as cyber aggression rather than using traditional bullying as a conceptual framework to define cyberbullying (Corcoran et al. 2015; Grigg 2010).

Even with the increasing number of studies that have explored the issues around defining and measuring both traditional and cyberbullying (Frisén et al. 2008; Hellström et al. 2015; Langos 2012; Naylor et al. 2006; Vaillancourt et al. 2008), further research remains important as inconsistencies may lead to errors in documenting prevalence rates or the efficacy of interventions (Modecki et al. 2014). There is substantial value in understanding how young people themselves define bullying, although this is still frequently ignored. Instead, emphasis is often placed on aligning young people’s definitions to better fit with academic definitions (Maunder et al. 2010). Yet, young people’s definitions should arguably inform scholarly definitions as it is during this life stage that bullying is most likely to occur (Carroll-Lind 2009; Monks et al. 2009). Effectively, by gaining deeper insight into young people’s own understanding of bullying, research may begin to eliminate some of the incongruences and inefficiencies in both research and practice in order to provide more efficacious and meaningful interventions.

The Current Study

Previous research has yet to resolve the gaps in defining bullying behaviours across research spheres, public perceptions, and personal experiences. Many studies assessing the incidence and outcomes of bullying present participants with a priori definition of bullying or ask them to respond to a questionnaire listing various bullying behaviours. Such methods are employed in order to provide clarification for behaviours that constitute bullying and thus produce more accurate and reliable data. However, these methods have also been found to restrict young people’s responses, and have been suggested not to represent their actual understanding or definitions of bullying (Canty et al. 2014). The present study, therefore, seeks to address this issue by eliciting young people’s views on what bullying means to them and the explanations behind their evaluations, without an a priori definition. As such, this study was carried out using a mixed methods methodology with the following aims: (1) to clarify how young people define bullying, (2) to examine the core components of their definitions of bullying, and (3) to ascertain whether academic definitions align with their understandings.

Method

Participants

Participants were recruited from local high schools and community youth groups by using research posters and snowballing. A total of 20 participants with a targeted gender balance (50% females) participated in the study, aged between 14 and 17 years old. In order to gain a diverse set of opinions and experiences, there was no restriction on whether participants had personally experienced bullying (in any capacity). The study was approved by a Psychological Human Ethics Committee under delegated authority by a University.

Development of Instruments

Scenarios

Four scenarios were developed by the researchers to each describe one of the major types of bullying: physical, verbal, exclusion, and cyberbullying (Rivers and Smith 1994; Slonje et al. 2013; Van der Wal et al. 2003). The physical setting of each scenario was also chosen based on previous evidence that it was a realistic and common bullying environment experienced by young people (Mishna et al. 2009; Monks et al. 2009; Olweus 1995). For the sake of brevity and to allow for greater discussion with participants, the scenarios were kept as short as possible, only consisting of a couple of sentences. For example, the exclusion scenario was based on the extra-curricular activity of a team sport: ‘Every week after the team’s soccer game, the players always organise to hang out later to relax and discuss the match. James/Jess never gets invited. When he/she tries to include him/herself and asks where they’re meeting, the other players always make up an excuse for him/her not to come’. No outcome or resolution was included in the scenario in order to provoke further discussion with the participants. Scenarios were presented to the participants in paper form and the order was randomised to control for response bias. The gender of the characters in the scenarios was changed to match the (visible) gender of the specific participant; thus, there were two sets of scenarios: one with female characters and one with male characters. This matching allowed for extra discussion regarding the perceptions of gender differences and prevented the data from pertaining to only one gender (for the purposes of the current study an extensive discussion of gender differences is outside the scope of the manuscript, although gender analyses have been undertaken).

Interview Schedule

To restrict the scope of the data collected, a structured interview schedule was developed based on the objectives of the study. The first section included questions that were asked after the presentation of each of the scenarios. The questions addressed (1) the perceived acceptability of the situation, (2) the perceived rationale for the behaviours, (3) who (if anyone) was harmed in the scenario and why, (4) whether the situation was considered to be bullying and why or why not, and (5) the influence of contextual factors. The second section of the interview schedule included two open-ended questions: (1) In your own words, can you describe what you think bullying is? (2) What do you think it would take to reduce bullying?

Procedure

Each interview was conducted in a location chosen by the participant. All participants were required to provide signed consent before the commencement of the interview, and participants under the age of 16 years were also required to provide signed consent from a parent or guardian. Participants were assured that their responses would be completely confidential, and that no identifying information would be included in the transcripts. Participants were presented with each scenario and asked a series of questions regarding the scenario, dictated by the interview schedule. After participants had answered all the questions to their own satisfaction, the next scenario was presented. This process was repeated for all four scenarios, and upon conclusion, the additional general questions were posed. After the interviews were completed, participants were given a debriefing sheet, which explained the research in greater depth and provided local support groups should the participants require further information or guidance. All interviews were conducted by the lead author. The length of the interviews was not restricted and was dependent on the participants’ length of response and discussion; interviews ranged in duration from 20 min to an hour and a half, resulting in approximately 20 h of recorded information in total.

Data Analysis

The interviews were audio recorded and transcribed verbatim using the Otranscribe software. The transcriptions were then entered into version 10.1 of QSR NVivo. Thematic analysis was used to analyse the data, following the set of broad of steps described by Braun and Clarke (2006). In order to gain a broad understanding of the data, the primary researcher read transcripts as a whole whilst making notes of possible themes and conceptual relationships. Known as familiarisation, this is the first important step which involves making sense of the text and enables the researcher to reflect on the overall meaning and gain a deeper understanding of data (Braun and Clarke 2006).

Thematic analysis seeks to identify themes emergent in the data, allowing for a more accurate depiction of the prevalent discourses. The next step, coding, is the systematic process of organising the data into relevant features and themes. It is the first step in identifying and grouping together patterns in the data. Both a deductive approach and an inductive approach were used in order to code the data. Specifically, deductive analysis was initially conducted by coding of the data by the specific interview questions as collapsed across the scenarios, as well as by the scenario as collapsed across questions, and then analysing these structural codes for meaning and interrelationships. Following this step, inductive analysis was carried out across the data corpus to search for emergent themes. The next stage of the data analysis consisted of interpreting and attaching meaning to the themes that were developed through both levels of analysis. This step was achieved through searching for patterns, regularities, and irregularities in the data. Here, a map of key patterns starts to develop (Braun and Clarke 2013; Clarke et al. 2015). Reviewing themes is the fourth step, ensuring that the coded data is organised in a consistent and coherent way. Each theme needs to be distinct and clearly adds to the overall picture. Defining and naming themes is carried out next, where the researcher summarises the conceptual meaning of each theme. The final stage is to then formally write up the data in an organised framework made up from the themes. Analytic conclusions should be made across all the themes (Braun and Clarke 2006; Clarke et al. 2015).

Results

The objective of the present study was to ascertain whether scholarly definitions of bullying behaviours accurately reflect how young people define bullying behaviour. One of the participants succinctly identified the problem currently faced by research:

I feel like almost everyone goes through it [bullying]. I don’t know if I’ve met anyone who hasn’t gone through or maybe they think they haven’t gone through it because no one knows what to define bullying as.

In order to address this gap, the following sections will highlight the perceptual differences held by young people regarding the academic components of bullying (harm, repetition, power, and intention) as well as introduce the concept of friendship in defining what constitutes bullying behaviour for young people.

Harm

All participants agreed that for a behaviour to be bullying, it had to be, by definition, harmful to the victim; that if the victim displayed distress, then the behaviour could be labelled as bullying. In fact, one participant suggested that harm was the only criteria for bullying:

I would say it’s [bullying] if you make someone else feel bad in any situation. No matter whether it’s a person, the media, a song, whether it’s anything, if someone ends up feeling bad because of something, that’s bullying.

However, most of the participants tended to emphasise the reaction of the victim in association with other contextual factors, although the victim’s reaction invariably had more influence in dictating whether the behaviour was bullying than the setting in which the behaviour occurred, for example, as one participant explained:

I think it all depends honestly how the person reacts to it. And that’s what changes the scenario, not necessarily whether it’s on social media or not.

It was apparent that when young people discussed bullying behaviours, they placed themselves in the scenario. In this way, they positioned themselves as the victim, and empathised with how it would feel. As such, the imagined reaction of victim and their personal pain were vital to their subsequent evaluation and judgement of the situation. Effectively, this determined the severity of the behaviour and whether it was considered as bullying or not. Given the extensive literature on the development of self and increased egocentrism present during adolescence (Sebastian et al. 2008; Somerville 2013), it is understandable that the personal experience of the victim would hold such importance for participants. In fact, previous research has also highlighted the importance of the victim’s experience in young people’s definitions of bullying (Frisén et al. 2008). Interestingly, participants did not often put themselves in the position of the aggressor, and thus tended to be less likely to have insight into the precipitating factors that would drive the bullying behaviours.

Repetition

Unlike harm, only half of the participants explicitly referenced repetition in their definitions of bullying. Three of these participants believed that bullying required repetitive behaviours, whereas the other seven argued against repetition, stating that bullying could be a one-off occurrence:

no it can just happen once, it doesn’t matter how many times it is. It’s bullying if you call someone a name once, it’s still going to make them feel bad about themselves even if you don’t do it anymore. Yeah you could call someone a bitch just one time and they’ll remember that for the rest of their life, that’s bullying. It can be anything.

These findings suggest that repetition is not necessarily fundamental to young people’s understanding of bullying, and some even actively suggest that it should not be included in our understanding of bullying at all. Given the strong focus on the reaction of the victim, it is not surprising that repetition did not feature within the definitional scope for most of the participants. Indeed, whether the behaviour occurs once or multiple times is redundant to the participants, as the victim has already experienced the behaviour and reacted in a way that indicates the behaviour is bullying. This is noteworthy as repetition of the bullying behaviours has been contested as a core component of bullying among researchers as well. In fact, as mentioned previously, the definition developed by the CDC suggests bullying can be a one-off occurrence if the behaviour is highly likely to happen again. This is in contrast to the participants in this study who indicated that bullying can be a one-off occurrence, but it is irrelevant if it is likely to reoccur, the key feature is the victim’s reaction when defining behaviours as bullying.

Power

At face value, youth also did not place any emphasis on the differences in power between the perpetrator and the victim. Only one participant explicitly discussed the importance of power imbalance between the bully and victim:

If it’s one on one, it’s normally a fight or argument but if one’s weaker than another, which most of the time means the stronger has more people than the weaker. They just use being strong as a weapon to hurt the weaker and they know that they’re doing that. I think getting hurt when you’re weaker than someone is bullying. If you get hurt with just one person with equal abilities, I think that’s not bullying. I think that’s a fight. Cause you have the ability to fight back.

Whilst Olweus (1993) stressed the existence of a power imbalance as central to the concept of bullying, it is possible that the current way of understanding power imbalances may be too narrow for young people to consider as relevant to bullying. Rather than constricting power to simply a ‘stronger vs weaker’ dynamic, young people may understand and wield power under a different guise. Although not explicit in their accounts, participants described underlying power dynamics through popularity and the publicity of interactions. For example,

I guess social media would be easier to break the person because it’s more in public, everyone sees it so it’s kinda like more embarrassing. And I think it would get more hurtful easily, just because the fact that everyone can see it and they’re seeing that and if they add on to it then it just makes it worse.

The participant described the heightened embarrassment and pain a victim might feel when bullied in public on social media. Here, the power dynamic is not maintained by a physical or emotional strength imbalance. Instead, the participant describes a relational type of power, where the victim has less power as they are the minority. Another young person also supported this theme, suggesting that bullies often made a scene to attract attention and cause more harm so that:

It’s more publicly embarrassing for the individual.

Consistent with these findings, Sticca and Perren (2013) found that young people perceived how public the bullying behaviour was to be a more important and detrimental aspect of bullying than the type of behaviour or setting in which the interaction occurred (Slonje and Smith 2008). It is plausible that for young people, the prospective embarrassment caused by a bully among one’s social group is where their power lies. Through public displays of bullying, either in person or over the internet, young people wield significant amounts of social power over potential victims. This is a key disparity between youth accounts of bullying and academic definitions, with research potentially failing to capture how power imbalances play out in young people’s daily lives. It is important to consider that teachers or adults in other supportive roles may in fact be missing possible harmful behaviour as their understandings of bullying do not include this subtle, relational power imbalance described by participants.

Intention

In the present study, the majority of participants asserted that behaviour does not have to be intentionally harmful to be bullying. One participant stated:

I don’t think it has to be intentional, cause again it could be subconscious and it’s the actions that end up hurting people.

Here, as per the prior analysis, the participant indicates that bullying is determined by the experience of the victim, not the motive of the bully. Another participant further clarified that intentionality was not considered to be a key factor because someone could feel bullied without the bully even realising it:

I think they could bully someone else without realising it. Definitely could do that. I think lots of people don’t realise they’re bullying someone. Like most of my friends probably wouldn’t have realised what they were doing was hurtful.

Again, the fundamental importance of the victim’s reaction is highlighted over other aspects of bullying. According to the young people in this study, even if the bully’s purpose of their action was entirely different, if the recipient was hurt in some capacity, then this was enough to define the action as bullying. These results are supported by previous findings where young people have not included intention as one of their requirements for bullying (Naylor et al. 2006; Vaillancourt et al. 2008). The following section on the emergent theme of friendship acts to elucidate the concept of intentionality.

Friendship

Of particular note in the current study, when asked to analyse a set of behaviours which are understood to reflect bullying, a common response was ‘it depends on their relationship’. The consensus among participants was that most behaviours that are commonly defined as bullying would not be bullying if the bully and victim were friends. Rather, the behaviour would be viewed as a joke between friends, as described by one participant ‘if they’re friends, it’d most likely just be seen as playful fun or whatever’. Yet it was later elaborated on by the same participants that friends can bully each other, but the intention to harm is not as prevalent as it would be among strangers. Thus, implying that bullying between friends is not to be considered as serious or as harmful:

Friends can bully each other at times but generally most of the time it’s banter and it doesn’t really mean anything because they don’t really mean it in a way. They just do it just for the time being and then they stop. Whereas if people weren’t friends, they would constantly do it more and more over time and that would be, I’d classify that as bullying.

Another participant extended this idea by suggesting that verbally aggressive behaviour between friends is not considered as bullying, but the behaviour may still trigger some negative thoughts and emotions in the victim:

It depends what the relationship is between the people. It would still be bullying, yes. But if the three of them were all friends, it would not be seen as bullying to (victimised character in the scenario). It would be the same situation where it would be like is this- are they having a laugh, are they serious kind of thing? Because it’s a hit and miss situation really, they can ask her but when they start laughing at her, you’re like are they laughing because I’m different? Or are they laughing because they hate it and they want me to change?

Linking back to the reaction of the victim, participants discussed that if the individuals were friends, the victim may not evaluate the behaviour in the same way and their level of tolerance for potentially harmful behaviour would be likely to be much higher.

In addition to the complexity of friendships and bullying, all participants described the role of joking within friendships and suggested that the line between joking and bullying behaviour can be blurred. According to participants, the distinction between joking and bullying often lies in whether both parties participate in the exchange. If one person does not reciprocate the behaviour and does not like it, then the behaviour becomes unacceptable:

does (the person) know it’s a joke? That’s how you know if it’s bullying. Cause if (they) know it’s a joke and maybe she’s joking back and then it’s fine but in that situation, she’s not joking back. I would say it’s affecting her.

This alludes to an aggressive dimension of humour, whereby an individual is being entertained at the expense of their peer (Martin et al. 2003). Participants understood that jokes can be hurtful and have a negative impact on the recipients, as described in the following quote

like I said before, it is bullying if it gets too far... there’s a really thin line between joking and bullying. So I think it’s like both. I think it’s like a joke and bullying because jokes still hurt.

This type of humour has been linked to hostility, low self-esteem, and negative emotions (Yip and Martin 2006). Once again, participants used an individualistic lens to evaluate the context and left it up to the recipient to determine whether the behaviour is a joke or not. That is, participants positioned themselves alongside the recipient, assessing the interaction from only the recipient’s individual perspective rather than taking both perspectives into account as a collective interaction between multiple individuals.

Through participants ongoing positioning with victim’s perspective, the importance of individual differences in levels of tolerance becomes apparent. An action that would be hurtful to one person may be considered ‘friendly banter’ to another. Participants were aware of the increased ambiguity this added when defining bullying behaviours.

Some people are more sensitive than others…But I get that everyone has a different line between the two, because everyone has different emotions and everyone’s been through different things so their capacity for that kind of banter is bigger than others. Yeah so I think it just depends on the person…like people joke around they think ‘oh I could take it, that’s not even mean’. But it actually is because the person, what they can handle is different than, to the other person.

Participants also identified that jokes are important to friendships because they allow individuals to acknowledge interpersonal issues without necessarily being hurtful. They understood that teasing can be good for individuals in certain contexts, for example it can make light of something that may otherwise be uncomfortable when discussed in a serious manner.

well if they’re your friend then you’re probably accepting them for your differences. If they’re truly your friends then I think you’ve already gotten over their difference and when you’re teasing you’re just sort of pointing them out. Saying you’re like that and I’m like this and I’m just going to point it out for whatever reason. I think joking is kind of important because it makes the subject, like you don’t have to approach it you can just leave it and everyone’s fine with it.

These findings highlight the complex and ambiguous nature of friendship as related to bullying. It is a key influence and possible mitigating factor for young people when determining whether an interaction is bullying or not. Irrespective of the positive functions that joking behaviour can serve in relationships, it is important to understand that jokes can be interpreted negatively and may be used to facilitate bullying behaviours.

General Discussion

The main objective of the current study was to explore the definition of bullying behaviours held by young people and whether academic definitions align with their experiences and understanding. Scholarly descriptions of bullying often include criteria that summarise the key features of bullying behaviours. It appears that to some degree young people also identify including harm, power, and intentionality in their definitions. Yet, these concepts are not necessarily functionally equivalent. There is a different understanding of what harm, power, and intention mean; the importance that they hold; and how they play out in friendships versus traditional bully-victim interactions. In fact, there are many more complexities and nuances to behaviour present in real-life situations which are intrinsically linked to how young people understand bullying. Specifically, in this study, young people considered the relationships between individuals at the core of how they describe bullying. Yet, interestingly, the onus was placed on the victim to take meaning from the behaviour, whether in a positive or negative way. It was the victim’s perception that was the most important: how they assessed the situation, how they reacted, and how they felt. This emotive individualistic perspective prevented participants from being able to separate features of the behaviour from the situational and relational context; as in fact, it was these contextual factors that dictated whether a behaviour was considered bullying.

Strengths and Limitations

Qualitative research has been known to impose a priori conventional definition of bullying on participants, which effectively primes participants’ responses to align with academic bullying criteria from the outset (Canty et al. 2014). As a result, any disparity between researcher and participant definitions of bullying is minimised and participants’ true interpretations are lost, thus obscuring the very phenomenon that research seeks to uncover (Canty et al. 2014). Accounting for this limitation was a particular strength of the present study. Although the four scenarios were developed by the researcher and thus illustrated a researcher-generated paradigm, a conventional definition of bullying was never presented to participants. In fact, participants were asked to describe what bullying meant to them in their own words. This allowed for disparities in bullying definitions to be identified, discussed, and critically analysed rather than labelled as inaccurate and redundant as has occurred in previous research (Canty et al. 2014; Vaillancourt et al. 2008).

The present study purposely took an approach that emphasised participants’ perspectives and allowed them the space to discuss the nuanced, intricate nature of relationships and situations in real life. The age range of participants was carefully considered to include the ages known to be the most likely to experience bullying, as it is these individuals who are arguably the most informed about how bullying is perceived, experienced, and dealt within in real-life contexts.

While this study has a number of strengths, there are nevertheless limitations that deserve discussion. One such limitation was that the four scenarios were short, with each only describing one particular behaviour. It could be argued that the present study failed to truly measure young people’s understanding of bullying behaviours as only a small set of potential examples were discussed. However, the research design encouraged participants to broadly discuss their definitions of bullying irrespective of the scenarios. As discussed above, the participants were not provided with any definition of bullying and were asked to generate their own, which enabled a much broader scope of behaviours and contexts to be elicited.

Furthermore, the scenarios only included interactions between characters of the same gender. It is difficult to conclusively study the effect of gender on bullying when the scenarios were same-sex interactions nested within the gender of the respondent. Thus, young people’s perceptions of mixed gendered bullying were not examined or compared with their perceptions of same-gender bullying. To overcome this limitation, the gender of the bully, victim, and respondent needs to be considered in future research.

Applications and Future Research

The present study has identified that certain components of bullying central to young people’s understanding are currently being overlooked by research. This in turn could have severe consequences for the interventions that use bullying research as their evidence base and may be an explanation for why the efficacy of interventions programmes has been called into question (Merrell et al. 2008). According to Merrell et al. (2008), the minority of interventions produce minimal positive effects, whereas the majority produced no effects at all. Rather than changing behaviours, interventions were more effective for creating awareness and changing attitudes.

Based on the current findings, it could be argued that interventions’ minimal success is due to an overemphasis on components deemed important by researchers (such as an individual’s intent to harm or the repetition of behaviours), and an underrepresentation of the bullying components important to young people. If interventions solely focus on behaviours outlined by academic definitions of bullying, they run the risk of being too narrow in their focus and targeting the wrong behaviours, which could be an explanation for a lack of positive change. In fact, the Scottish Government has taken the stance that all bullying behaviours are contextually bound and considers bullying to be ‘both a behaviour and (an) impact’ where all incidents must be assessed individually (Scottish Government 2017). Specifically, they indicate that both the behaviour and the outcomes must be treated in ways that address the intentionality and impacts in sensitive ways that reduce the likelihood of unacceptable and hurtful behaviours. This conceptualisation overcomes some of the definitional issues which seek to objectively define bullying behaviours, and inadvertently ignore the subjectivity inherent within peer interactions.

It is suggested that future research and interventions should not only base their programmes on how young people define bullying but also focus on the important social factors that have been shown to influence young people’s perceptions and subsequent behaviour. In line with Stevens et al.’s (2001) recommendation, interventions should also target parents and the family as well as young people to maximise behavioural changes in multiple environments. Intensive programmes which incorporate the social context of bullying, such as including parental engagement, have been found to be the most effective (Ttofi and Farrington 2010). Such programmes may assist in educating and empowering those who are not directly involved to identify who is experiencing bullying and how they are being impacted. It must be noted that a fundamental component of all interventions is to identify the areas most in need, raise awareness, and provide preventative and coping strategies (Sticca and Perren 2013). Young people’s knowledge, as demonstrated in the present study, should truly be at the core of bullying interventions, as well as used to inform future initiatives and raise awareness about how seemingly harmless peer interactions which may have been overlooked using objective measures of bullying could have prolonged significant impacts on individuals.