Abstract

Keywords: Risk awareness; online risks; children; qualitative methods; cross-cultural research

Introduction

In this article we aim to shed more light on the social construction processes whereby children and young people make sense of online risks, by investigating the role of media representations in shaping children’s perceptions of online negative experiences.

Important results have been achieved through empirical research on the reported risky experiences that children encounter online (Livingstone, Haddon, Görzig, & Ólafsson, 2011; Mascheroni, & Ólafsson, 2014). Less is known, however, about the influence of public and media discourses on risk awareness and perception among younger internet users. Indeed, prior research that analysed media coverage of internet risks and harm pointed to the impact of media representations on the public agenda and on the process of social reflexivity (Haddon & Stald, 2009, p. 390). Nonetheless, the main focus of these analyses was on adult recipients of media discourses, including parents, teachers and other educators. Moreover, analysis showed that children have little if no voice in the media coverage of online risks: they are represented mainly as passive agents and spoken for by social institutions (Ponte, Bauwens, & Mascheroni, 2009).

By contrast, we claim for a child-centred approach to online risks and safety (Livingstone et al., 2011). As Davis postulates, qualitative research in the contexts within which young people’s experiences are situated and moulded helps uncover “a variety of children’s voices” (1998, p. 327). Within this framework, reconstructing children’s own discourses about risks, and how they appropriate and negotiate media representations, enables a deeper understanding of spaces and cultures that adults do not belong to. Staksrud and Livingstone (2009) define online risk as the “heterogeneous set of intended and unintended experiences which increase the likelihood of harm to the internet user” (p. 4). Harm can be conceptualized as the distinct outcome of exposure to risk (Livingstone, 2013). Yet, not every experience of risk results in harm (Livingstone & Haddon, 2012). This implies that the co-construction of online risks is of particular interest for understanding children’s online experiences, as it is on the basis of what is socially constructed as problematic that children define, negotiate and adopt preventive measures aimed at reducing unpleasant and harmful consequences.

In an attempt to tackle the gap in our understanding of children’s own perceptions of risks, we draw on rich qualitative data where children and young people define online risk and discuss their internet-related experiences in their own terms, from a cross-national perspective. We focus on the discourses of children and young people that allow us to understand the social processes whereby risk perception is constructed, highlighting the role of media, parents and peers.

How Children Make Sense of Risks

Whereas realistic perspectives posit risk as an objective process (Beck, 1992), by contrast, constructionist approaches have emphasized the importance of contextual and subjective factors in the definition of what constitutes risk (Lupton, 1999). Deborah Lupton argues that the perception of what is risky or not varies not only according to different historical, cultural and social settings, but also based on different individual characteristics, and specific situations. This doesn’t mean that audiences reject public representations of expert knowledge. Rather, people often make “a self-consciously calculative and strategic” use of experts (Lewis, 2006, p. 468), and exhibit modes of reflexivity about risks that reflect their own cultural and economic capitals (Threadgold & Nilan, 2009). Moreover, ordinary audiences construct their own definition of risks – including online risks – in often quite ambivalent ways. In fact, lay people negotiate the discourses they gather from public and private spheres, drawing on an ever-changing morality (Bauman, 1993). The pleasure that can be involved in risk-taking also needs to be taken into account when we try to understand ordinary people’s risk perception and (sometimes voluntary) experiences (Tulloch & Lupton, 2003). In discussing young people and risk, a constructionist perspective is particularly fruitful for understanding their experiences, including risk-taking ones, in the context of the relations of power among different children, and between children, parents and other adults (Ponte, Simões & Jorge, 2013).

Moreover, looking beyond actions and behaviours, and into attitudes and perceptions, it is crucial to understand how ordinary audiences position themselves vis-à-vis expert knowledge (Beck, 1992). The concept of sensemaking of risk, taken by Wall and Olofsson, consists of “the way people materialize meanings of risk within a social context” (2008, p. 432). Therefore, this concept situates individual understandings of risk as reflecting social contexts and particular social and cultural situations within them (Threadgold & Nilan, 2009).

The perception of risk by young people is thus influenced by their social and cultural condition, with all that entangles of interaction with parents, peers, school and the media. Media and peer cultures have gained relevance in the cultural socialization of children, while parents and teachers have lost their authoritative role, claims Pasquier (2005). It is thus often among peers that the boundaries of what is perceived as risky, acceptable or correct are established, regarding others, be it friends, acquaintances or strangers (Flores & James, 2013; Hagen & Jorge, forthcoming). Joke Bauwens describes how, by questioning the dominant morality transmitted by parents and establishing a new context-specific set of rules for online behaviour, young people create a “peer-driven morality” (2012, p. 44). Part of this morality is also established via the online participatory culture where children and young people interact more and more, which poses precisely the challenge that children become part of this online culture while being exposed to alternative norms (Jenkins, 2009).

Understanding the ambivalent positioning children take regarding dominant discourses on what risks consist of, helps explain why children may act differently online according to different contexts, but also complicates the line experts draw between risks and opportunities. Meeting strangers online tends to be portrayed as one of the most dangerous risks of the internet, but is also perceived by children and young people as an opportunity to meet new people and to sometimes develop romantic relationships (Barbovschi, Marinescu, Velicu, & Laszlo, 2012; Ito, 2009). A cross-national study may, then, reveal not only different perspectives on what constitutes risk, but also, as we wish to discuss in this article, the different processes whereby the perception of online risk is constructed.

Media Influence on Individual, Parents’ and Peers’ Perceptions of Online Risks

Within media studies, the role of the media in shaping public discourses has been a recurrent and pervasive field of investigation. The tradition of agenda setting studies, for example, identifies media effects – understood in terms of long-term, cognitive effects – in the ability of news coverage not only to set the agenda of issues of public concern, but also their power to “frame” social discourse, that is, to provide audiences with certain perspectives and interpretations of these very issues (McCombs, 2005; Weaver, 2007). Focusing more on fictional media content, cultivation theory has shown how media shape audiences’ perceptions of risks, by over-emphasizing certain risks while downplaying others (Gerbner, Gross, Morgan, & Signorielli, 1986; Gerbner & Gross, 1976).

From a sociological and socio-psychological perspective, the question has been reframed in terms of how media representations contribute both to the social construction of reality (Berger & Luckmann, 1966) by setting the boundaries of reality itself, and to sediment and convey social representations (Farr & Moscovici, 1984). In other words, in representing reality, the media embody social discourses and turn them into publicly accessible, popular symbolic resources: media representations mirror lay discourses, while at the same time informing and shaping common-sense knowledge. Evidence of the co-construction of reality on the part of both media and society is offered by the study of moral panics: by defining what counts as a social problem, the media are simultaneously “an expression of panics, as well as their spark or cause” (Goode & Ben-Yehuda, 2010, p. 90). However, media representations – and moral panics – are then incorporated into individual experiences and collectively made sense of, on the basis individuals and social groups’ systems of meaning (Thompson, 1995).

The co-determining of media and social representations should be understood against the backdrop of the pervasive symbolic power of media in the so-called “reflexive modernity” (Beck, 1992; Giddens, 1990, 1991), characterized by an increasing “epistemological suspicion” (van Zoonen, 2012) and a decline in trust in expert knowledge as well as in the media as a knowledge institution. In this perspective, the internet, and social media more specifically, amplify the turn into the self, personal feelings and experiences as the source of true and reliable knowledge, that van Zoonen (2012) labelled “I-pistemology”. The emphasis on the self as an authoritative position to speak from does not, however, undermine the pervasiveness of media representations in lay discourses: while the claim of truth advanced by the media is clearly contested, social actors still rely on media to gather knowledge, and combine media representations with personal experiences and word of mouth. As van Zoonen puts it, “a combination of media information, personal experience and popular wisdom underlie the way people make sense of events in the world around them” (2012, p. 62).

When it comes to children and media use, the role of the media in influencing the public debate (Haddon & Stald, 2009), also through the diffusion of “media panics” (Drotner, 1999), is well documented. The media raised a variety of online risks on the agenda: popular news stories in recent years include episodes of cyberbullying, sexting, “stranger danger” and pornography (Haddon & Stald, 2009; Ponte et al., 2009). The analysis of media representations of online risks demonstrates that news coverage of these phenomena is affected by sensationalism, and that the media’s constructions of childhood and children’s online experiences influence how these issues are framed within national cultures (Ponte et al., 2009). Prior research has also shown that cross-national variations in the media attention for online risks are mirrored in parental concerns around the internet. This can also be explained by the fact that parents turn to the media as one of the primary sources of information on e-safety: in 2008, a Eurobarometer survey (European Commission, 2008) showed that broadcast media were ranked as the second most important source of information on safer internet use, after closest family members (Ponte & Simões, 2009). Similarly, in the EU Kids Online survey, parents indicated traditional media as the second most common source of information after friends and family (Livingstone et al., 2011). Also, when asked where they would like to get advice on online safety, parents included the media as a preferred source, after school (Staksrud & Ólafsson, 2013). We can thus conclude that media representations influence the public, policy and research agenda, if not directly, then by shaping the discursive contexts in which parents, teachers, policy-makers and scholars are dealing with the social issue of children and the internet (Haddon & Stald, 2009).

As anticipated, however, little is known about how children incorporate media discourses in their own perceptions of online risks, and whether media influence is direct or in what ways is it mediated by parents or peers. To bridge this gap, this contribution aims to: (1) understand how and to what extent children’s perceptions of online risks incorporate media representations, parental worries and discourses circulating among peers (e.g. news that are then appropriated and form part of the peer group’s common-sense knowledge); (2) identify any age- or gender-specific patterns in the appropriation and conversion of media, parents’ and peers’ discourses; and (3) identify cross-cultural variations in risk perceptions among young people (e.g. concern for different risks in different countries) that might relate to differences in the media agendas (Ponte et al., 2009) or differences between countries (Helsper, Kalmus, Hasebrink, Sagvari, & de Haan, 2013; Livingstone et al., 2011).

Methods

To reach this goal we draw on the empirical material collected within the EU Kids Online III network in nine countries (Belgium, Czech Republic, Greece, Malta, Italy, Portugal, Romania, Spain, and the UK). Fieldwork was carried out from February to September 2013, with varying degrees of difficulties due to different procedures of consent to gain access in the schools in each country (see Barbovschi, Green, & Vandoninck, 2013, pp. 60-77). On average, six focus groups (three with girls and three with boys, with two groups for each age group of 9-10, 11-13 and 14-16) and 12 interviews (six for each gender, with the same age distribution) were conducted in each country. Overall, 56 focus groups (N = 254) and 114 interviews (N = 114) took place across the nine countries. Though systematically taking into account socio-demographic differences is beyond the scope of the present analysis, children were recruited in both public and private schools, and in different areas (urban, suburban and rural) (see Table 1; see also Smahel & Wright, 2014).

Table 1. Overview of data collection in the nine European countries (EU Kids Online III).

Focus groups and interviews were transcribed in national languages and coded based on a coding manual developed by the coordinators and tested by each national team: at least two transcripts were independently coded by two researchers from the same country, in order to guarantee reliability and quality of coding. The first level of coding (Rennie, Phillips, & Quartaro, 1988) resulted in a rich amount of condensed descriptions of the material in English, including information on the situation, actors involved, and any further comment the researcher deemed important for the understanding of the interview. The most relevant excerpts were also translated into English.

First-level codes were then thematically analysed through a second level of coding done by a different group of researchers, based on a coding guide which distinguished: (1) the general area (risk experience, risk impact, risk perceptions, preventive measures, online activities, mediation, literacy, opportunities, researchers’ comments, off-topic); (2) the problematic situation (strangers, bullying and harassment, sexual content and contact, unwanted content, commercial risks, personal data misuse, technical problems, health and overuse); (3): the platform (SNS, email, pop-ups, websites, chats and messages, video platforms, games and virtual worlds, online phone and video, school platforms, boards and forums, mobiles and tablets); (4) the actors involved (respondent, friends and peer group, siblings, parents, teacher and school, the media, other known people); and (5) any explicit negative or positive feeling mentioned by respondents (yes/no). The reliability of coding across coders was ensured, with a minimum inter-coder reliability of 0.70 (Kappa) for each category.

The software NVivo was then used to sort out children’s discourses that were coded as “risk perceptions”. Given the particular interest in the incorporation of media representations within the process of sensemaking of risks, the material coded as coming from “the media” and made a search by keywords (newspapers, news, television, MTV, documentary, stories, etc.) was selected. Overall, we analysed 5,176 first-level codes.

Results and Discussion

For each of the risks analysed in the second level of coding, we identified the main components of children’s discursive repertoires (definition of risks; the vocabularies that children draw on in making sense of online negative experiences; ideal-type situations; actors involved; responsibilities; imagined harmful consequences of risks), as well as the main sources of awareness. Overall, five main sources shape children’s perceptions of what is problematic on the internet: first-hand online experiences; second-hand or indirect – at times, even mediated – experience that is part of peer culture, circulating via word of mouth; parental concerns, which are expressed in forms of advice as well as restrictions; awareness-raising and e-safety programmes in school, for example, online safety initiatives run by the police; and media representations, which influence children’s knowledge and attitudes both directly and indirectly (via peers and parents). For the purpose of the present article, we focused mainly on media discourses, stories by peers that have a source in news, and parental concerns informed by media debate and media panics.

Moreover, we selectively focused on two online risks: cyberbullying and “stranger danger”, because their distinctive features help us to isolate the appropriation of media representations within children’s discourses. Indeed, these risks are likely to mobilize diverse discursive frameworks as they refer to different relational contexts: while cyberbullying and harassment are contextualized within peer-to-peer relationships, “stranger danger” is associated with a communicative context in which the child is positioned as the recipient of a communication initiated by others (Hasebrink, Livingstone, Haddon, & Ólafsson, 2009). Therefore, we assume that social representations of cyberbullying rely more on narratives, vocabularies and anecdotes circulating among peers via word of mouth, while “stranger danger” is socially constructed as an online risk through a deeper adhesion to media frames and language. Moreover, we hypothesize age, gender and country variations in the discursive repertoires, whereby younger children, who have a limited online experience, are more vulnerable to media representations, while teenagers co-construct online risks through engagement with peers – and, therefore, differences in the process of the self-epistemology and self-reflexivity of risk (Lewis, 2006; van Zoonen, 2012).

“Stranger Danger”

“Stranger danger” – especially in the form of online sexual solicitation and cybergrooming – is a key component of media panics around children and online communication (Ponte et al., 2009), which are then mirrored in adult and children’s concerns around the internet. Indeed, together with bullying, strangers top the list of risks that children perceive as the most dangerous, scary and worrisome for its potential offline harmful consequences: as a 10-year-old boy in a focus group in the Czech Republic claims, commenting on the group’s construction of strangers as the main danger of the internet, “the strangers will hurt you physically.”

Against these lay perceptions of the dangers associated with online contact with people never met face-to-face before, empirical evidence suggests that “meeting strangers” can encompass a variety of circumstances and experiences, which cannot be assumed as universally problematic (Barbovschi et al., 2012; Wolak, Finkelhor & Mitchell, 2008); indeed, evidence shows that many children interact online with unknown others with little risk (Wolak et al., 2008, p. 343). The EU Kids Online survey showed that, while communicating online with someone never met face-to-face is the most common risky experience, characteristic of 30% of 9- to 16-year-old European children, a minority (9%) has an offline meeting with online contacts (Livingstone et al., 2011), and meetings are likely to occur mainly among peers. More importantly, the risk of being harmed from a face-to-face contact with someone met online is very low (Livingstone et al., 2011). Since the probability that children and their peers have directly experienced a harmful offline meeting with an online contact, we can thus assume that children’s perceptions of “stranger danger” are socially constructed based on media representations and parents’ concerns, which are appropriated and negotiated within peer cultures.

Children’s discourses on strangers. While children, especially teenagers, experience contact with people they have never met before as a clear opportunity, when talking about “stranger danger” they tend to distance themselves from their own positive experiences and adopt the media’s “moral panic” frame: they use a vivid, brutal vocabulary that evokes the sensationalist tone of most news coverage of crime stories (Ponte et al., 2009), as the following excerpt illustrates:

Or else he could take advantage of you before killing you. If you do not do what he says, he could touch your private body, hurt you, throw stones at you or smack you with a belt or stab you with a knife. (Malta, 9-10, girls)

Other recurrent terms in children’s vocabulary include “pederasts”, “paedophiles”, “perverts”, “rape” and “rapist”, and “kidnap”. In making sense of the “stranger”, children also adhere to the stereotypical representation of strangers as older men affected by mental disorders: in children’s own words, strangers are often depicted as “schizophrenic” or “crazy guys” who groom children online in order to later kidnap, rape, or even kill them.

Though the boundaries between contact with strangers and other online risks such as bullying, sexting and personal data misuse are not easy to draw – in children’s perceptions these experiences are likely to overlap – we can identify three typical situations and actors emerging from discourses on strangers collected in focus groups and interviews – grooming and paedophiles; fake identities; and thieves – and two negative outcomes of risky contacts and encounters that provoke children’s fears and worries – threats to an individual’s own body and security, or threats to one’s possessions and belongings.

Discourses on the “paedophile” suggest that children adhere to the stereotypes circulating in adults’ and media representations: hence, they socially construct “stranger danger” as a gendered risk, whereby the groomer is a male adult, predominantly middle-aged, to whom girls are more vulnerable. In making sense of grooming, therefore, children reproduce common-sense narratives in which grooming is coupled with a further ingredient of media panics: the belief that the technological affordances of online communication encourage anonymity and fake identities. Consequently, the groomer is framed as an older man pretending to be a young boy in order to contact girls, to persuade them to meet offline and to abuse them. The following excerpt provides an example of popular representations of groomers:

On Facebook you might accept people that you think are certain people, but then it turns out to be... different. My dad, for example, told me that the daughter of a colleague added a boy, they started talking, and they decided to meet. Ehm... but her mother accompanied her, because she has never met this boy before, and when they went to the meeting they could not see this guy. So she called him and the mobile phone of an adult nearby started ringing[...] they hide behind boys’ profiles so as to contact other young people. (Italy, 15, girl)

Fake identities and people pretending to be someone else were often debated in focus groups and interviews: children lament that online identities can never be fully reliable because people can easily use a fake picture and claim an incorrect age.

“Stranger danger” is also articulated in another common frame, one which associates strangers with the misuse of personal information and stealing of personal possessions: younger children in particular are aware that including their address on social networking site (SNS) profiles, as well as sharing pictures of their homes and updating plans for holidays, may be risky because strangers may be able to locate the house and steal things.

What influences children’s discourses? In making sense of “stranger danger”, children mobilize knowledge and narratives gained from the media, since first-hand experience of being groomed by an adult or being burgled because of personal information disclosed on the internet is fortunately limited. Examples like the following, in which children directly link their perceptions on meeting strangers with news coverage and media representations, are common across countries and age groups:

The influence of media representations on children’s perceptions is not restricted to news and crime stories: the popular MTV’s reality show, Catfish, is also mentioned by interviewees and focus group participants in Belgium, Italy, Portugal, Malta and Spain as an important source of information on “stranger danger”. Others refer to programmes for children that raise awareness about online risks, as explained by one boy:

There was a programme about how there was a man and there was a little girl and the man sent a picture of himself[...] And he said he’d got tickets to some sort of band or play or something. And then she got there with her mum, and her mum was waiting in the car for her, because he only had two tickets, and it was a fully-grown man and she got really scared and stuff, and didn’t know what to do. So, sort of warning you about it, but in a kids’… A TV programme, yes. Just about the dangers of the internet. (UK, 9, boy)

Media representations also inform risk perceptions in a more indirect way, by shaping the discursive environment in which the process of sensemaking of risks is embedded (Wall & Olofsson, 2008): the borders of media influence are expanded and amplified by parents’ and peers’ discourses. The following excerpt from a focus group with 13-year-old girls in Belgium is emblematic of the process through which news stories are incorporated into lay discourses and inform both adults’ and peers’ experience. Moreover, it also signals how news stories that have a global resonance – in this case, the story of a British girl having her house destroyed after promoting a party on Facebook – are locally appropriated and made familiar through association with similar stories circulated by word of mouth:

Children therefore re-assemble and personalize stories gathered from the media in order to make them closer to their every day life contexts, and hence, more reliable and authoritative (van Zoonen, 2012).

Age and gender differences. Despite perceiving contact with people never met face to face as also an opportunity, when discussing risks, boys and girls of all ages were equally likely to adhere to stereotypical representations of “stranger danger” as a middle-aged groomer who pretends to be someone else, or a burglar. This discourse is strongly gendered. While the belief that girls are more vulnerable to strangers is part of lay discourses, it is susceptible to alternative framings. More specifically, in Italy, Spain, and partially, in Greece, girls who have contact with strangers on the internet are perceived not only as victims, but also as co-responsible for harmful consequences: adding strangers on Facebook because a high number of friends is a marker of popularity; posting provocative pictures or spending too much time on Facebook are all risky behaviours that expose girls to stranger danger:

Then you see young girls who go out with... and also adults who write ‘Hi sweetie, let’s meet offline’, and so. And then these girls accept to meet them and are raped – I mean sexual abuse is never right, but I claim: what are you doing yourself? Why did you accept to meet a stranger who contacted you on Facebook? (Italy, 14-16, girls)

Boys share this peer-driven morality, though girls might also conform to it. The reference to girls’ sexual reputation is an indicator that a sexual double standard is at play here: as other research has shown, while sexting has become “a new norm of feminine desirability” (Ringrose, Harvey, Gill, & Livingstone, 2013, p. 319), girls who choose to conform to this normative practice are later called “sluts”.

Cyberbullying

Cyberbullying has been defined as intentional, repetitive aggression involving power imbalance between the victim and perpetrators, which is carried out by means of electronic communication (Juvonen & Gross, 2008; Schrock & boyd, 2008). In a recent contribution, Marwick and boyd (2014) draw a distinction between bullying and other forms of online harassment labelled as “drama”, where the latter is reciprocal, while the former is grounded in power imbalances. Other scholars have also pointed to different degrees of cyber aggression, noting, however, that victims of less serious online harassment still engage in active coping responses - e.g. distancing themselves from the event and reframing it as non serious (Machackova, Cerna, Sevcikova, Dedkova, & Daneback, 2013). So despite victims of cyberbullying tend to adopt different strategies, the study suggests that online harassment can also have harmful consequences that children try to cope with.

Indeed, in their discourses on cyberbullying, children refer to a variety of perceptions and experiences, including witnessing, experiencing or being aware of hate speech, disrespect, fights, gossip, rude and nasty comments, annoying, unwelcome or sexual messages or ones that are outright cyberbullying, that happen mostly on SNS. Children are also concerned about data misuse, offences, threats, blackmail, insults, trolling, when they encounter swearing (more specifically when playing games) and to a lesser extent, racism or religious harassment.

Albeit not an insignificant percentage, only 6% of the children surveyed in the EU Kids Online II project claim to have been bullied on the internet; the same data show, however, that cyberbullying is the most bothering and harmful experience children can encounter online (Livingstone et al., 2011; Livingstone, Kirwil, Ponte, & Staksrud, 2013). Indeed, even though most of the interviewees had never experienced cyberbullying themselves, children associate what they perceive as bullying or harassment with a wide range of negative emotions and bad consequences. Similarly, when imagining the impacts of online harassment, children are aware of painful consequences, including the most serious such as depression and even suicide.

Children’s discourses about bullying and harassment. Children draw on a language of extremes when it comes to referring to cyberbullying, using terms such as “horrible” (Czech Republic, 12, girl), “pure evil” (Portugal, 12, girl) or describing this experience as something which can “crush you” (Spain, girl, 12). Since only a minority of the interviewees have experienced any form of cyberbullying, this language is part of the shared discursive repertoire children draw upon, and could be attributed to how the media often portray stories of cyberbullying that have dramatic consequences, such as suicide.

Several interviewees and focus group participants mentioned cyberbullying as being the “worst” thing that could happen online. A deeper analysis of children’s discourse about bullying and harassment, however, confirms the heterogeneous, multifaceted nature of children’s perception of this online risk.

In almost all the countries involved in the study, children distinguish between jokes or teasing, nasty comments and bullying. Children themselves admit they partake in jokes and teasing, and how some nasty comments are not always bullying: “having fun” and “making fun of” (Greece, 15, girl) are distinct. In line with what Marwick and boyd (2014) found, children often express how some fights on Facebook are fun to watch, especially since there are no offline consequences. Moreover, there seems to be a shared understanding that if the incident is isolated and not repeated over time, it does not necessarily belong to the category of cyberbullying. By contrast, online harassment is classified as bullying when the act is more repetitive, ongoing and pervasive, and often carried out by more than one person. This distinction makes it possible for young people to frame their experiences of online conflict in distinct ways from the bullying narratives adopted by adults (Marwick & boyd, 2014, p.11). Indeed, children do not seem comfortable labelling whatever form of conflict on the internet as bullying; they recognize that sometimes it can be a misinterpretation, a mistake or a joke.

However, children are also aware that the line between teasing and bullying is fine, and what can start as a joke might easily escalate into bullying, with severe consequences. The distinction between teasing and bullying is often related to the source. Children perceive that sometimes what comes from their friends can be teasing, so they mention taking different measures for dealing with friends and strangers. They find it easier to report strangers who harass them, but would think about it twice if it was a friend. This mirrors the concept of sensemaking of risk (Wall & Olofsson, 2008), and how the understanding of risk is context-sensitive (Hagen & Jorge, forthcoming).

The language of extremes is also matched by a language of absolutes that the children seem to use when referring to cyberbullying. Their discourse includes notions of how “they [referring to online bullies] are always coming up with new things” (Malta, 14, girl), and that “there will always be bullies” (Belgium, 14, boy). Discourses around cyberbullying reveal a sense of helplessness and the belief that cyberbullying cannot be controlled per se. Lack of control is further specified as the difficulty, on the part of the victim, of subtracting oneself from a situation of bullying. With this awareness, children themselves seem to be reflecting that other solutions, rather than the somewhat impossible task of eliminating bullying, are to be sought. Children suggest to their peers that they should block, report and try to avoid places such as Ask.fm, where the possibility of encountering bullying is higher, and avoid having any material (such as sexts) that they could be harassed with.

The feeling of powerlessness against online aggression is mirrored in the perceived differences between face-to-face bullying and cyberbullying: the online environment is perceived as a space where one can be less cowardly and more courageous in expressing oneself, but also more inclined to fight and insult others. In line with what has been highlighted by studies on cyberbullying (Levy, Cortesi, Crowley, Beaton, Casey, & Nolan, 2012), interviewees associate online anonymity as the one afforded by Ask.fm, with an intensification of online threats, as they could say hurtful and painful things from a safe distance, or what some children refer to as being “behind the screen”.

What influences children’s discourses? Children’s discourses about cyberbullying are a good indication that awareness-raising initiatives are achieving the intended effect. However, it is also clear that children’s discourses are also influenced by what the media frame as cyberbullying. Children as young as 10 mention knowing of, or having heard of, deaths and suicides associated with cyberbullying. Some children explicitly identify the media as the source of information. This can be assumed as an indicator of the way globalized media representations of cyberbullying shape and frame children’s discourses about cyberbullying. The children often mentioned the case of Amanda Todd¹. In certain countries, such as Malta, even though there were no local deaths associated with cyberbullying, Maltese children still mentioned suicide as a potential consequence of cyberbullying. As Critcher (2003) argues, the media convey stories of cyberbullying and give voice to the moral panic associated with online harassment, while simultaneously causing further panic to arise.

The media exercise both a direct and indirect influence on children’s perceptions, through the mediation of parents. And parents seem to use the media as a gateway to discuss the dangers of cyberbullying with their children.

In the case of bullying, however, children’s own and their peers’ experiences are also important sources of information. Children are in a “privileged position to speak from” (van Zoonen, 2012, p.60): sharing experiences promotes informal learning through which children acquire the necessary tools to react and cope with negative experiences online. When compared with other sources if information, first-hand or peers’ experiences provide children with better resilience and know-how to deal with such matters. This finding is consistent with prior empirical evidence that a greater exposure to risks is correlated with more resilience (Livingstone et al., 2011).

Age and gender differences. Older children seem to have a higher awareness of the persistence of digital information, acts of cyberbullying included. In contrast, younger boys express a feeling of powerlessness against anonymous cyberbullying, which might indicate a lack of awareness of reporting mechanisms. Consistent with prior research on cyberbullying (see Livingstone et al., 2011), older children are seemingly more able to cope with insults and feel less bothered, while younger children’s discourses suggest higher vulnerability and helplessness. Increased awareness by teenagers might be attributed to greater internet experience and higher exposure to online risks, as well as better media-savviness that could also be the result of listening to news stories in the media, discussing them with peers and families, leading them to be more sensible online.

Age differences in coping with online harassment and drama are also visible in the vocabularies that children of different ages adopt. Teenagers express very harsh judgements of peers’ online behaviour, for example, labelling sexting as “silly” and bullying on Ask.fm as “stupid”. As anticipated, victims of sexting are often blamed as responsible, though senders are also recognized as co-responsible.

When it comes to gender, both girls and boys seem to agree in their perception that girls have more fights online, are more dangerous, but are also more vulnerable than boys (see also Marwick & boyd, 2014). However, children themselves debate whether how one reacts to online conflicts and aggression is more a question of one’s personality than one’s gender.

Conclusions

In this article we have examined how children make sense of online risks; more specifically, we were interested in assessing how the discursive environment produced by media representations and adult’s constructions of the dangers of the internet shapes the way these risks are perceived by children.

Findings show that media framings of online risks are particularly influential when children are less likely to experience those risks directly: in making sense of the “stranger danger” children adhere to stereotypes, vocabularies and frames that they draw from the media (both news media and entertainment or educational TV formats). Similarly, younger children are more vulnerable to media representations of the most serious consequences of being bullied online. By contrast, teenagers who boast stronger experience of the patterns and risks of online conflicts draw on first-hand or other peers’ experiences in making sense of cyberbullying. However, beyond exerting a direct influence on children’s perceptions, media narratives are also incorporated in lay discourses that circulate in peer exchanges as well as within the parent–child relationship. This re-embedding of media discourses in the context of everyday life often entails a personalization and localization of media coverage, which turns distant stories into personal and realistic experiences. Through a similar re-embedding, media representations are re-positioned as true experiences of peers, and thus reliable sources of knowledge, in compliance with the norms of “I-pistemology” (van Zoonen, 2012).

Against evidence of cross-national variations in how the media cover online risks (Haddon & Stald, 2009; Ponte et al., 2009), cross-cultural comparison did not point to striking differences in the sensemaking of risks in the countries involved in this study – one exception being the highly gendered definition of both “stranger danger” and bullying that result from sexting, whereby girls in Southern European countries are particularly negatively sanctioned as responsible for being involved in such risky situations. One reason for homogeneous perceptions across countries lies in the global character of the media environment children grow up with: some crime stories related to internet risks have a global resonance, thus reducing variations on the national media agendas. A common sensationalist tone adopted by different media outlets, including the quality press (Ponte et al., 2009), in the framing of both meeting strangers and cyberbullying, further reduce cultural differences: the dominant discourse of fear and the portrayal of the child as a victim operate to the detriment of alternative, empowering discourses on online risks and opportunities for children. Moreover, TV formats such as Catfish are broadcast in many countries and become taken-for-granted components of youth cultures in a variety of contexts.

In terms of policy implications, then, media framing of online risks might pose a challenge to awareness-raising and education campaigns: children are more often exposed to stories that might be instilling fear rather than empowering them to take action and to stand up for themselves. At the same time, the re-use of media narratives in children’s sensemaking of risks suggests that children are more likely to remember and learn from real-life stories: including examples from peers who children perceive as close to their own experiences within education campaigns could be useful to help children further their knowledge about online risky interactions, and to develop efficient preventive and coping strategies.

A second reason for low cross-cultural variations, though, can be identified in what is also the main limitation of the present study: nine countries from the EU Kids Online network volunteered to participate in this research. Thus, not all the types of countries classified by EU Kids Online on the basis of the opportunities and risks of the internet for children and parental mediation strategies (Helsper et al., 2013) are represented here. Similarities in the framing of online risks may also be the outcome of homogeneous social contexts and comparable online experiences, because we also found similar age patterns among respondents regarding the awareness of risks. Future research with children in countries characterized by different cultures of childhood and parenthood as well as different patterns of children’s online experiences will shed more light on the processes of social construction of risks and children’s negotiations of adults’ and the media’s discourses.

A further limitation of this study lies in the challenges of cross-cultural comparison of qualitative data. As Livingstone recognizes, “these data are context dependent, reliant on the local knowledge of the researcher, and their transcripts are even written in different languages” (2003, p. 488). As in other studies, several methodological challenges were found and dealt with in designing and implementing this qualitative study. Strategies for controlling the context of the research included ensuring the highest homogeneity as possible in the recruitment of children from schools, and conducting focus groups and interviews in the school setting with only the presence of the researcher. However, as anticipated, cultural differences regarding ethical aspects of doing research with children as well as different legal and formal requirements to gain access to school and the children emerged (Barbovschi et al., 2013).

Various strategies for minimizing the power imbalance between the researcher and children participating in the research were adopted (Barbovschi et al., 2013). Nonetheless, we are aware that interviews are socially constructed situations distinguished by a set of scripts, norms and expectations (Potter, 2004): we cannot exclude, therefore, that children have conformed to reciprocal expectations on what is socially legitimate and desirable.

However, we were interested in exploring how children’s perceptions are discursively constructed, also drawing on media discourses, and discourses are not immune from normative expectations and stereotypes.

Acknowledgements

The EU Kids Online network is funded by the Safer Internet Programme of the European Commission. We would also like to acknowledge COST Action IS0906, Transforming Audiences, Transforming Societies, for supporting the collaboration for this article through a Short-Term Scientific Mission.

Notes

1. Amanda Todd was a Canadian teenager who committed suicide in October 2012, after posting a video on YouTube in which she reported her experience of being bullied. Amanda was persuaded by a stranger to show her breasts on camera. The picture was circulated online, causing strong emotional distress for the young girl.

References

Correspondence to:
Giovanna Mascheroni
Department of Sociology
Università Cattolica del Sacro Cuore
Largo Gemelli 1, 20123 Milano

Email: giovanna.mascheroni(at)unicatt.it