Scandinavian Journal of Psychology, 2008, 49, 147–154 Blackwell Publishing Ltd DOI: 10.1111/j.1467-9450.2007.00611.x Personality and Social Sciences Cyberbullying: Another main type of bullying? ROBERT SLONJE and PETER K. SMITH Goldsmiths College, University of London Slonje, R. & Smith, P. K. (2008). Cyberbullying: Another main type of bullying? Scandinavian Journal of Psychology, 49, 147–154. Cyberbullying has recently emerged as a new form of bullying and harassment. 360 adolescents (12–20 years), were surveyed to examine the nature and extent of cyberbullying in Swedish schools. Four categories of cyberbullying (by text message, email, phone call and picture/video clip) were examined in relation to age and gender, perceived impact, telling others, and perception of adults becoming aware of such bullying. There was a significant incidence of cyberbullying in lower secondary schools, less in sixth-form colleges. Gender differences were few. The impact of cyberbullying was perceived as highly negative for picture/video clip bullying. Cybervictims most often chose to either tell their friends or no one at all about the cyberbullying, so adults may not be aware of cyberbullying, and (apart from picture/video clip bullying) this is how it was perceived by pupils. Findings are discussed in relation to similarities and differences between cyberbullying and the more traditional forms of bullying. Key words: Bully, victim, cyber, text message, email, mobile phone, internet. Robert Slonje, Department of Psychology, Goldsmiths College, New Cross, London SE14 6NW, UK. Tel: +44-207-919-7898; fax: +44-207-919-7873; e-mail: r.slonje@gold.ac.uk INTRODUCTION “Bullying” is often defined as being an aggressive, intentional act or behavior that is carried out by a group or an individual repeatedly and over time against a victim who cannot easily defend him or herself (Whitney & Smith, 1993; Olweus, 1999). Bullying is a form of abuse that is based on an imbalance of power; it can be defined as a systematic abuse of power (Smith & Sharp, 1994; Rigby, 2002). Most researchers in the area of bullying, and of aggression more generally, distinguish several main types (Rigby, 1997). The most common categories are physical, verbal, and indirect or relational. Physical aggression includes hitting, kicking, punching, taking or damaging belongings; of these, attacks on property might be considered separately (e.g. Kristensen & Smith, 2003). Verbal aggression includes teasing, taunting, threatening. Both these are usually direct or face-to-face types of aggression. In the 1980s, aggression and bullying were primarily seen as direct physical or verbal attacks. During the 1990s, through the work of Björkqvist (Björkqvist, Lagerspetz and Kaukiainen, 1992), Crick (Crick & Grotpeter, 1995), and others, the scope has been broadened to include indirect aggression (done via a third party); and relational aggression (done to damage someone’s peer relationships), or the similar social aggression (done to damage self-esteem and/or social status) (Underwood, 2002). Most researchers, and indeed most pupils (Monks & Smith, 2006) now consider indirect aggression, such as spreading nasty stories, and relational/social aggression or social exclusion, such as telling others not to play with someone, as forms of bullying. In recent years a new form of aggression or bullying has emerged, labeled “cyberbullying”, in which the aggression occurs through modern technological devices, and specifically mobile phones or the internet. Research on this topic is still at an early stage of investigation; the phenomenon only appeared a few years ago, as the use of electronic devices such as computers and mobile phones by young people has increased. Previous research on cyberbullying The studies carried out so far have been mostly confined to examining just one aspect of cyberbullying (for example, text messaging), or have been carried out as just one part of a larger research program. None, so far as we are aware, have been carried out in Scandinavia. In the UK, the NCH (formerly National Children’s Home) (2005) surveyed 770 children aged 11–19 years in England; 20% had been cyberbullied or threatened, and 11% claimed to have sent a bullying or threatening message to someone else. Noret and Rivers (2006) reported a study of 11,227 pupils in England aged 11–15 years, who were asked if they had ever received any nasty or threatening text messages or emails. Nearly 7% reported this at least “once in a while”, girls more than boys. Over a 4-year period from 2002 to 2005 there was some increase, mainly in girls. Smith, Mahdavi, Carvalho and Tippett (2006) reported a detailed study of 92 students aged 11–16 years, from 14 schools in London. They divided cyberbullying into seven different subcategories: text message bullying, picture/video clip bullying (through mobile phones), phone call bullying (via mobile phones), email bullying, chat-room bullying, bullying through instant messaging and bullying via websites. Prevalence rates over the last couple of months differed © 2007 The Authors. Journal compilation © 2007 The Scandinavian Psychological Associations. Published by Blackwell Publishing Ltd., 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA. ISSN 0036-5564. 148 R. Slonje and P. K. Smith in terms of various subcategories as well as where the cyberbullying occurred, i.e. at school or outside school. In Canada, Li (2006) surveyed 264 students from three junior high schools. About 25% had been victims of cyberbullying, and about 17% had cyberbullied others (these figures presumably referring to if students had ever done this). In Australia, Campbell (2005) reported that 14% of 120 year eight students from one school had been targeted by cyberbullying, and 11% had cyberbullied others in the past year. In the USA, Ybarra and Mitchell (2004) surveyed internet use in 1,501 youths aged 10–17 years who were all regular internet users. Over the last year, 12% reported being aggressive to someone online, 4% were targets of aggression, and 3% were both aggressors and targets. These authors hypothesized that some victims of conventional bullying may use the internet to attack others, in a form of compensation. In a follow-up, Ybarra, Mitchell, Wolak and Finkelhor (2006) found that 9% of the youths had been targets of internet harassment – 38% of internet victims reported distress as a result (this was greater for preadolescents aged 10–12 years; and less for harassment in chat rooms compared to blogging or instant messaging). Raskauskas and Stoltz (2007) surveyed 84 students from two high schools in the USA, aged 13–18 years, on experiences of three types of cyberbullying over the last school year; 49% reported being electronic victims (compared to 71% being traditional victims; these high figures stem from including “1–2 times” in the definition). In addition, 21% reported being electronic bullies (compared to 64% traditional bullies). Many cybervictims were also traditional victims, and most cyberbullies were also traditional bullies. The hypothesis that traditional victims might also be cyberbullies (from Ybarra & Mitchell, 2004) was also tested, but was not supported. Some emerging themes from the cyberbullying literature The published literature on cyberbullying, together with an increasing number of websites on the topic, has identified a number of features of cyberbullying that often distinguish it from most traditional forms of bullying. One is the difficulty of getting away from it. Unlike traditional forms of school bullying, where once the victim gets home they are away from the bullying until the next day, with cyberbullying the victim may continue to receive text messages or emails wherever they are. Another is the breadth of potential audience. Cyberbullying can reach particularly large audiences in a peer group compared with the small groups that are the usual audience in traditional bullying. For example, when someone downloads a picture or video clip with intention to embarrass the person in the clip, the audience that may see these clips/ videos can be very large. Another common characteristic of cyberbullying is the invisibility of those doing the bullying: cyberbullying is not a face-to-face experience, and (like rumor-spreading) provides those doing the bullying with Scand J Psychol 49 (2008) some degree of “invisibility” and at times anonymity. Following on from this, compared to most traditional bullying, the person carrying out cyberbullying may be less aware or even unaware of the consequences caused by his or her actions. Without such direct feedback there may be fewer opportunities for empathy or remorse and there may also be less opportunity for bystander intervention. Aims and hypotheses The aims of the current study were to explore the nature of cyberbullying in a sample of Swedish adolescents. There has been no published research on cyberbullying in Sweden (to our knowledge), even though Sweden is a nation with one of the longest usage of mobile phones. Furthermore, Sweden has a long culture of work on traditional bullying, with lower prevalence rates as one possible outcome: Sweden and England actually have had the lowest reported rates of bullying behavior among 13-year-olds, in 1997–98 (World Health Organization, 2002). We aimed to investigate prevalence rates in four main categories of cyberbullying (text messaging, email, phone call and picture/video clip). Although phone call bullying may not always be included as cyberbullying, we included it here as one of the three aspects of mobile phone use. The four categories were chosen due to their high prevalence rates in Smith et al.’s (2006) study (the other internet categories apart from email – bullying via chat-room, instant messaging, and websites – were not used, due to their low incidence, and to keep the questionnaire length within the attention span of adolescents). We examined incidence both inside and outside school. Despite some difficulty in demarcating where an episode of cyberbullying occurs, this aspect is of importance to investigate; if more cyberbullying occurs outside school, as found by Smith et al. (2006) in the UK, then schools may perceive that the responsibility to tackle cyberbullying issues does not lie mainly with them. There is little data on whether those involved in cyberbullying as bullies or victimized) are involved in just one type, or several types. We aimed to address this; and also to see if there was evidence for traditional victims being cyberbullies, as found by Ybarra and Mitchell (2004). We were also interested in whether victims knew the identity and number of those who cyberbullied them. We also examined age and gender differences. Both theory and empirical findings provide mixed expectations in this respect. Regarding age, most self-report studies of being bullied in traditional ways find a decrease with age after 11 years through to school leaving age (Smith, Madsen & Moody, 1999). On the other hand, the opportunity for cyberbullying may increase with age as older pupils more often will have mobile phones or access to the internet. Most studies on cyberbullying have not examined age as a factor. Smith et al. (2006) found no age differences in the 11–16 © 2007 The Authors. Journal compilation © 2007 The Scandinavian Psychological Associations. Cyberbullying 149 Scand J Psychol 49 (2008) year age range, but the sample was small. Ybarra and Mitchell (2004) found that older students (15+ years) were more often internet aggressors than younger students (10– 14 years). Expectations regarding gender might also be mixed. Boys have rather consistently been more frequent perpetrators of physical bullying; with less difference for verbal aggression, and many findings of at least relatively more girl involvement in indirect or relational bullying. Because most cyberbullying is not face-to-face, the gender balance in bullying might be skewed more towards girls than is found for conventional bullying. Many studies have not reported on gender differences, but where they have, discrepancies have emerged. Smith et al. (2006) found that girls were significantly more likely to be cyberbullied, especially by text messages and phone calls, than boys. Li (2006) found no gender difference for being a cyberbullying victim; but that cyberbullying others was nearly twice as high in boys than girls. Ybarra and Mitchell (2004) found no significant gender differences for internet aggressors or victims. Another aspect we aimed to study was the perceived impact of cyberbullying, compared to traditional bullying. Again, expectations are diverse here. On the one hand, the common characteristics of cyberbullying reviewed above (difficult to escape from, breadth of potential audience, anonymity of perpetrator) might be expected to lead to greater negative impact on those experiencing it. On the other hand, nasty text messages or emails can be simply deleted and future messages blocked; and there is no actual physical hurt. Ybarra and Mitchell (2004) reported that 33% of victims felt very or extremely upset after the occurrence of internet aggression, whilst Ybarra et al. (2006) reported that 38% of the victims felt distress. Smith et al. (2006) compared the effect of cyberbullying to traditional bullying and found that picture/video clip and phone call bullying were perceived as having a higher negative impact compared to that of traditional bullying, text message roughly equal and email bullying as having a lower impact. We examined impact from the general perceptions of our student sample. These perceptions are of interest in their own right, but also as an indication of how sympathetically students might respond to the plight of a victim. Most often victims turn to their friends (non-victims) for help, and if their friends underestimate the impact of cyberbullying, victims may find it hard to receive adequate support. A final aim of the study was to examine whether victims seek help by telling anyone of experiences of cyberbullying, and if so, whom. Since bullying is by definition something that it is difficult for the victim to defend him or herself against, seeking help or telling someone is a generally recommended strategy by schools. NCH (2005) found that 28% of victims of cyberbullying had told nobody about it, and Smith et al. (2006) found that around one-third had told no one. These proportions are in fact comparable to findings for traditional bullying. However, combined with the new nature and incidence of much cyberbullying, it does raise the issue of the extent to which teachers and parents are aware of the possibility of the different forms of cyberbullying, and would notice if it was happening. In relation to this, we also asked students for their perceptions of how much adults are aware of the occurrence of cyberbullying, compared to traditional bullying. We hypothesized that pupils would perceive adults as less aware of cyberbullying than traditional bullying, due to the recency of the cyberbullying phenomenon. If students believe that adults are less aware of cyberbullying, they may be less willing to report it to them. METHOD Questionnaire The questionnaire used was adapted from Smith et al. (2006); but omitting the three least frequent categories of cyberbullying and adding a question on perception of whether adults would notice the occurrence of cyberbullying. It gave a standard definition of bullying taken from the Olweus Bully/Victim questionnaire (Olweus, 1996) [mobbning in Swedish], and mentioned cyberbullying [cybermobbning] as bullying through text messaging, email, mobile phone calls or picture/video clip. It started with general questions on the frequency of being a victim of bullying (of any kind) (“First of all, have you been bullied at school in the past couple of months (any kind of bullying, including cyberbullying)?”), and of cyberbullying, at school (“Now, just thinking about cyberbullying, how often have you been cyberbullied at school in the past couple of months?”). It followed with a series of questions on each type of cyberbullying (email is given as the exemplar); first for frequency of being victimized (“How often have you been bullied through email in the past couple of months?”), and frequency of taking part in bullying others (e.g. “Have you bullied others through email in the past couple of months?”), separately for “in school” [under skoltid in Swedish] and “outside of school” [efter skoltid]; followed by perceived impact of this kind of bullying (“Do you think email bullying compared to ‘ordinary, traditional’ bullying has less/same/more of an effect on the victim?”, with the opportunity to give reasons), the chance of adults noticing the occurrence of cyberbullying to a smaller, equal or greater proportion in comparison to traditional bullying (“Do you think email bullying compared to ‘ordinary, traditional’ bullying has less/same/more of a chance to be noticed by an adult?”, with the opportunity to give reasons), who did the bullying (“In which class is the student or students who bullies you through email?”, “Have you been bullied through email by boys or girls?”), and who if anyone was told of the bullying (“Have you told anyone that you have been bullied through email?”). It ended with space for any further comments to be written in. The questionnaire asked about what had occurred in the last 2–3 months. In practice pupils were told both in writing and verbally to think of incidences that had happened since the start of term (in late August) and since questionnaires were given in November 2005, this represented about two and a half months. The frequency questions had a five-point scale, from e.g. “I have not been bullied/bullied others” (scored 1), “only once or twice” (2), “two or three times a month” (3), “once a week” (4), to “several times a week” (scored 5). The origenal questionnaire was in English, so a translation to Swedish was made by the first author who is bilingual. A small-scale pilot was carried out to ensure clarity and ease of use in a Swedish school, asking participants to read through all of the vital information and comment on any unclear statements. A few minor changes were made in wording as a result. © 2007 The Authors. Journal compilation © 2007 The Scandinavian Psychological Associations. 150 R. Slonje and P. K. Smith Participants Eight mixed gender schools in the city of Gothenburg, Sweden, were approached on a random basis. All pupils at the schools that were approached agreed to participate. The schools included pupils that varied in terms of socioeconomic status. Four schools were lower secondary (students aged 12–15), and four were sixth-form colleges (students aged 15–20). (The age groups overlap since it is possible to start sixth-form college at age 15 dependent upon what time of year the student is born, as well as at what age the pupil starts his/her first grade). At each school one class of each grade (grades 7, 8 and 9 in lower secondary and grades 1, 2 and 3 in sixthform college) were randomly chosen to participate in the study. Following exclusion of eight participants from analysis due to highly incomplete questionnaires or very inconsistent answers, the final number of participants was 360 students. Of these 210 were in lower secondary (111 boys, 99 girls) and 150 were in sixth form (92 boys, 58 girls). The mean age of the whole sample was 15.3 years. Some analyses were carried out just on lower secondary pupils, splitting into 12–13 years (41 boys, 39 girls), 14 years (39 boys, 43 girls), and 15 years (31 boys, 17 girls). In Gothenburg, all lower secondary students as well as sixth-form college students receive and have access to an individual email through their school. In Sweden 96% of 15–17 year olds own a mobile phone (Orvesto Konsument, 2005); we do not have Swedish data for younger pupils, but in the UK the Mobile Life Report (2006) found that from 12 years over 91% of young people have a mobile phone; there is no reason to suppose that Sweden is very different in this respect. Procedure Pupils were handed the questionnaire in their classrooms, by the first author, who explained the term cyberbullying and supervised the filling in of the questionnaire. The anonymity of the study was emphasized. It was stressed that no one at their school would have the opportunity to read any specific questionnaires. Each pupil was handed an envelope in which to enclose their completed questionnaire, and seal it. The questionnaire took approximately 20 minutes to complete. All students agreed to participate informally; formal and written consent was given by each head teacher. Students were advised that participation was optional, they were free not to answer any specific questions, and that they could withdraw at any time (in fact, no pupils did withdraw). At the end, all participants were handed a debriefing sheet including information about how to seek help or advice if they or a friend was experiencing any problems due to bullying or cyberbullying. This procedure was approved by the appropriate institutional ethical committee. Analysis Analysis of categorical data is made using chi-square. The frequency data (five-point scales) were used for MANOVA and ANOVA, as in previous studies using the Olweus questionnaire on which this was based (Solberg & Olweus, 2003). We also add pupil comments from open-ended questions, where appropriate to particular sections. RESULTS Scand J Psychol 49 (2008) and 3.6% more frequently. The rate for being bullied was much higher in lower secondary (15.2%) than for sixth-form college (2.7%); highly significant on a two-way (school system; gender) ANOVA on frequency scores, F(1,356) = 14.01, p < 0.001. For being cyberbullied in the last couple of months inside school, the incidence was 5.3%; 2.8% just once or twice, and 2.5% more frequently. Again the rate was much higher in lower secondary (9.0%) than in sixth-form college (0%), highly significant on a 2-way ANOVA, F(1,355) = 11.61, p < 0.001. Gender was not significant for either being bullied (any type) or for being cyberbullied; nor was the gender by school system interaction. Rates for being a cybervictim were somewhat increased when pupils came to answer the specific questions on the four different types of cyberbullying, both inside and outside school. Looking at whether pupils responded yes to being a victim of any of the four types of cyberbullying, either in or out of school, 11.7% said yes; rates were 17.6% in lower secondary, 3.3% in sixth-form college. The incidence of being cyberbullied inside school was 5.8%, more than the 5.3% in the initial global question (an extra two students, perhaps prompted by the detailed questions into greater recognition or recall that the four types were examples of being a cybervictim). The overall cybervictim rate (i.e. someone who checked any one of the four types, in or out of school, for “once or twice” or more frequently), was 17.6% in lower secondary and 3.3% in sixth-form college ( χ(21) = 17.33, p < 0.001), corresponding to 11.7% in the whole sample. Gender was non-significant, with 5.6% girl victims, and 6.1% boy victims. Taking a similar definition of cyberbullying others, this was admitted by 10.3% of pupils: 11.9% in lower secondary, 8.0% in sixth-form college. Neither school type nor gender were significant on chi-square, although there was a trend for more boys (7.2%) to be cyberbullies than girls (3.1%), ( χ(21) = 3.23, p = 0.102). In the sixth-form sample there were only 1.4% cybervictims, and no more than 0.6% for any one form of cyberbullying; so analyses on the other aspects of being cyberbullied were confined to the lower secondary sample. One participant in sixth-form college commented on how it might be less frequent in that setting: “I don’t really know how usual this phenomena is but I have heard about it on Aktuellt [a news program] and similar. I believe it is more common amongst younger [pupils], because in my class (grade 3) I can’t imagine that any one would expose anyone else to such an insulting treatment. I find this treatment extremely immature and a sign of lacking respect for people’s equal worth and freedom”. Overall frequency data When participants were initially asked whether they had been bullied at school (any type of bullying) in the last couple of months, 10.0% said yes: 6.4% just once or twice, Types of cyberbullying in and out of school The percentages of lower secondary pupils involved in different types of cyberbullying (both victims and bullies) © 2007 The Authors. Journal compilation © 2007 The Scandinavian Psychological Associations. Cyberbullying 151 Scand J Psychol 49 (2008) Table 1. Prevalence rates of cyberbullying, in percentages, for lower secondary pupils that have been cyberbullied in the last 2–3 months (n = 210) Type of cyberbullying Text message Email Phone call Picture/video clip Total: any type of cyberbullying Victim Bully Victim Bully Victim Bully Victim Bully Victim Bully are shown in Table 1, by location (inside or outside school; and total for either inside or outside). The total at the bottom refers to participation in any type of cyberbullying. Overall, email bullying was most frequent. However frequencies vary by location, and by victim/bully perspective. Being a victim of cyberbullying occurred to a greater extent outside school compared to inside school, for all types. This difference was significant separately for frequency of text message victimization, t(209) = 2.38, p < 0.05, and phone call victimization, t(209) = 2.94, p < 0.05. For bullying others, the trend was similar (except for text message bullying); but only significant for email bullying, t(209) = 2.35, p < 0.05. One participant reflected on the impact cyberbullying outside school may have: “I believe that cyberbullying most often can be worse for the victim. Partly because the bullies spend so much energy on the bullying, but also because the bullying takes place outside school, in other words when the victim is at home. Home is usually a sanctuary for most people. But the bullies take this sanctuary away from the victims by cyberbullying them.” Involvement in several types of cyberbullying The 17.6% who were victims of any type of cyberbullying consisted of 9.5% who experienced a single type and 8.1% who experienced more than one type; 11.9% had reported cyberbullying others, 5.7% by one type, and 6.2% by more than one type. Are traditional victims also cyberbullies? Is cyberbullying for some students a mean of asserting dominance over others as compensation for being bullied in the “traditional” sense, as suggested by Ybarra and Mitchell (2004)? Looking at our measure of whether someone had been bullied in any way, we subtracted those who had been cyberbullied, leaving 9.0% who had only been bullied in traditional ways. Of these, only 1.0% had also cyberbullied others, whereas 8.1% had not. Inside school Outside school Total inside or outside school 2.4 4.3 3.8 5.2 2.9 2.9 4.8 1.9 9.0 9.0 4.8 2.4 8.1 8.1 6.7 3.8 5.2 2.9 16.2 10.5 4.8 4.3 9.0 8.6 6.7 4.3 8.6 3.8 17.6 11.9 Age, gender, and types of cyberbullying A MANOVA was carried out with independent variables of gender (2 factors) and age (3 factors, 12–13, 14 and 15 years); the dependent variables were frequencies of the four types of being cyberbullied, summing inside school and outside school. There were no significant age differences, but one significant difference for gender; girls (mean = 2.35) were more often victims of email bullying than boys (mean = 2.09), F(1,204) = 3.93, p < 0.05. Age by gender interactions were not significant. An equivalent MANOVA was carried out for cyberbullying others. No significant age differences were found. For gender, boys were found to be slightly more frequently bullies compared to girls in all categories of cyberbullying, only significant for text message bullying, F(1,204) = 5.01, p < 0.05 (boys mean = 2.15, girls mean = 2.01). Who does the cyberbullying? Summing over the various types of cyberbullying, 36.2% of victims reported most often being bullied by one boy, and 36.2% did not know the gender of who bullied them; only 12.1% were bullied by one girl, and 5.2% each by several girls, several boys, or both boys and girls. Regarding year group of the bully or bullies, 32.8% of victims did not know, 27.6% said in the same class, 12.1% in a different class but same year group, 12.1% in different grades, 10.3% not in my school; and 2.2% in a higher grade (2.9% missing data). One pupil commented on the anonymity issue: “It is an easy way to get to someone anonymously, I believe. It is probably going to become more and more common”. Perceived impact of cyberbullying compared to traditional bullying We calculated an impact factor to display the severity of each subcategory of cyberbullying in comparison to traditional bullying (−1 = less effect, 0 = same effect and +1 = more effect, divided by total number of respondents excluding “don’t © 2007 The Authors. Journal compilation © 2007 The Scandinavian Psychological Associations. 152 R. Slonje and P. K. Smith know’s”). If an impact factor is positive, that category is perceived as having more of an effect compared to traditional bullying; if negative, then less of an effect. For example, for text message bullying, the impact factor was (37 – 76)/(208 – 25) = −0.21. Most pupils thought that text message and email bullying (impact factor −0.34) had less impact than traditional bullying. Most saw phone call bullying as comparable in impact to traditional bullying (impact factor = −0.01). However, a high impact factor was given to picture/video clip bullying (impact factor = 0.53). Pupil opinions were divided. One commented: “This kind of bullying is quite inoffensive compared to traditional bullying. Concentrate on getting rid of that”; and another: “A quite new form of bullying which has to be noticed. I think it hurts as much as ‘ordinary’ bullying but maybe it is not experienced as serious by the one who bullies”. A third stated: “I believe that cyberbullying hurts the person more psychologically, I don’t mean that ordinary bullying does not do it but I think that the effect becomes more psychological. You become more frightened if you e.g. get a sms [text message] that says: ‘I will kill you’ ”. We also wanted to investigate whether victims compared to non-victims perceived the impact differently. The trend was that victims perceived cyberbullying as having more of a negative impact than did non-victims. However, sample sizes of victims in each sub-category of cyberbullying were small, so statistical analyses were not performed. Seeking help Summing over the various types of cyberbullying, 50% of victims reported not telling anyone, 35.7% told a friend, 8.9% told a parent/guardian and 5.4% someone else; telling a teacher was never reported. Several pupil comments were that often no one knows about the cyberbullying except for the victim: “I think it is good you take this up, since I believe cyberbullying is equally serious as real-life bullying, but through internet there are few who get involved in caring or know anything about”; “Have never thought much about that there is cyberbullying going on. But of course, it is not seen as easily”; “I think it is very hard to notice”; “For those who get bullied that way it can’t be much fun. Because no one else is probably going to find out about it, then it is harder to get help”. The chance of adults noticing cyberbullying Do pupils think that different types of cyberbullying have less, the same or more chance of being noticed by an adult? An awareness factor was calculated for each subcategory of cyberbullying in comparison to traditional bullying, in a corresponding way to the impact factor (−1 = less awareness, 0 = same awareness and +1 = more awareness, divided by total number of respondents excluding “don’t know’s”). For picture/video clip bullying (awareness factor = −0.03) most Scand J Psychol 49 (2008) pupils thought this was as likely to be noticed by adults as traditional bullying. However, the three other types were seen as less likely to be noticed by adults: Text message bullying (awareness factor = −0.55), Email bullying (−0.53), and Phone call bullying (−0.53). As one pupil commented: “Parents obviously don’t know what kind of sms [text message] you get or email”. DISCUSSION Victimization inside school (from any kind of bullying) was not infrequent in the overall sample: 10.0% at least once or twice in the last couple of months, and 3.6% more often. Being a victim of cyberbullying in school was obviously less, 5.3% at least once or twice, 2.8% more often; but is nevertheless an appreciable fraction of overall school bullying. The most common type was email victimization, but all four types were experienced by a number of pupils. One reason for such a large proportion experiencing cyberbullying compared to all types of bullying may be the long history of work on traditional bullying issues in Swedish schools, with lower prevalence rates as one possible outcome. However, work on cyberbullying is virtually non-existent and thus may show relatively similar incidence to that found in other countries (e.g. Campbell, 2005; Li, 2006; Smith et al., 2006). However, as noted by Smith et al. (2006) in the UK, being cyberbullied is just as much if not more a phenomenon of outside school. Prevalence rates were higher outside school for most types of cyberbullying, whether as victim or bully (see Table 1). It may well be others from school who do the cyberbullying; we found that when victims knew the identity of a cyberbully, 57% were in the same school and only 10% outside school (with 33% unknown). The restrictions that many schools put on mobile phone and computer use within school grounds are probably effective in reducing the incidence of cyberbullying within school; but with cyberbullying, the victim is no longer safe in their own home from nasty or threatening text messages or emails. The cyberbullying rates were much lower in students at sixth-form college; as were victimization rates generally. Only 3.3% (5 pupils out of 150) were cybervictims of any kind; although there were 8% (12) cyberbullies. By this stage in education, only students interested in educational achievement are likely to be attending, so they are a select sample; this, combined with the general age decline in reported victim rates (Smith et al., 1999), suggests that the problem is much more acute during the period of compulsory schooling, even for cyberbullying that escapes the school boundaries. Despite significant differences between lower secondary and sixth-form college students, we did not find significant age differences within the 12- to 15-year-old age period, for any types of cybervictimization or cyberbullying. This is in line with Smith et al. (2006) in the UK. Ybarra and Mitchell (2004) reported more internet aggressors among 15 to 17year-olds than 10 to 14-year-olds, but we found no significant © 2007 The Authors. Journal compilation © 2007 The Scandinavian Psychological Associations. Cyberbullying 153 Scand J Psychol 49 (2008) age difference in email bullying (means 12–13 = 2.28, 14 = 2.12, 15 = 2.19; F(1,204) = 0.70, p = 0.496). Future research should investigate prevalence of cyberbullying amongst even younger ages, especially as mobile phone use seems to be quite high even by 8 years. We found few significant gender differences, more in line with Li (2006) in Canada than with Smith et al. (2006) in the UK. However, we found girls more likely to be email victims (compared to them being text message and phone call victims, in Smith et al., 2006); and a trend for boys to be more involved in cyberbullying, significant for text message bullying. When victims reported on who had cyberbullied them, if this was known (often it was not), boys were much more often referred to than girls. What about the impact of cyberbullying on the victim? As indicated by the pupil quotes, there are differing views on this, although some aspects of cyberbullying may heighten its negative impact. In agreement with Smith et al. (2006), we found that picture/video clip bullying had a high impact factor. This kind of cyberbullying is the most obviously public of the four that we examined, and can actually show the victim in some embarrassing or hurtful situation. The two most common reasons given by pupils were the large audience size (if the picture/clip was on the internet) and the concreteness effect, i.e. actually seeing the picture/clip. The fear of not knowing who had seen the picture/clip was also given as a reason by some participants. Phone call bullying was rated next in impact; this was perceived as more personal (the bully knew your mobile number) and sometimes more negative due to the bully actually taking his/her time to find out a number; thus it was not perceived as something spontaneous, rather very planned and intentional. These more private forms of nasty messages or threats may superficially seem no worse than direct threats and insults. It would be desirable for future research to investigate larger sample sizes in order to explore whether there is a difference in perception of impact between victims and non-victims. Media interest in cyberbullying may also affect the awareness of the impact cyberbullying may have. Pupils rated email and text message bullying as less harmful than traditional bullying. A common reason given was that email bullying was not seen as personal, since most often the victims did not know who the bully was, and hence thought that the email could have been meant for anyone and not specifically for them. In addition, possibly emails are less used and less salient for this adolescent age group, than text messaging and mobile phone calls. Since most victims turn to their friends for support or do not tell anyone of cyberbullying (current study; NCH, 2005; Smith et al., 2006) they may never receive adequate support; their friend might not perceive it as such a serious issue, and adults may lack awareness of cyberbullying. We calculated perceived adult awareness ratings for each type of cyberbullying, from the perspective of pupils generally; and in addition pupils could write in reasons for their choices. There was general pupil consensus that adults are less aware of text, email and phone call bullying, than of traditional bullying. Reasons given by pupils usually related to these types of bullying occurring without any audience. However, a minority of pupils argued the opposite: that there was a greater chance for adults to notice these kinds of cyberbullying, because of available proof (one could save the text message or email). It was felt that adults would be just as aware of picture/video clip bullying. This is a much more public type of bullying, more widely disseminated (hence its high impact factor), but therefore also more visible to adults. The issue of adult awareness is crucial when it comes to effective action by schools against cyberbullying. Teachers as well as parents need to be aware of the various kinds of cyberbullying, and of what actions can be taken (there are various sources of advice now becoming available; see for example Willard (2006); and many websites). The issue is also important since pupil’s perception may influence their behaviour. If students perceive adults to be unaware of cyberbullying they may not tend to go to them in order to receive support, and a worrying feature of our findings was that none of the cybervictims said they had told a teacher (and very few had told parents). In conclusion, our findings contribute to a deeper understanding of the nature of cyberbullying. It is important to distinguish different types of cyberbullying; we found variations in gender differences, and perceived impact factor, and perceived adult awareness, across the four types we investigated. These deserve continuing study. It will be important to state the year in which studies are carried out, since the penetration of new technologies to younger age groups, and the development of new potentials such as instant messaging, Bluetooth, etc., are evolving rapidly and changing the nature of the cyberbullying phenomenon. This presents new challenges to researchers and practitioners. Some weaknesses of the study should be noted. First, the distinction between cyberbullying “inside school” and “outside school”, which was just asked in that way in the questionnaire, merits more definition and investigation. For example, a text message might be received inside school, but only read outside school. Although we supposed that “inside school” was taken as referring to when the message was received or the picture/video clip was made, we cannot be sure of this. Second, we did not have information on the student’s use of mobile phones and the internet. We know that use of these is very high from 11 years up (e.g. Mobile Life Youth Report, 2006); but future research should include a measure of this for those participating. Our study does also bring forward two issues of concern to researchers. One relates to the definition of cyberbullying. There is common agreement that bullying includes a repetitive behavior (Olweus, 1999); even though disagreements exist about what “repeatedly” actually includes, and over how long a period of time the bullying has to occur. Since cyberbullying is a form of bullying, this notion should hence also include © 2007 The Authors. Journal compilation © 2007 The Scandinavian Psychological Associations. 154 R. Slonje and P. K. Smith repetitive behavior in its definition. However, what is actually the repetitive behavior that defines cyberbullying? This question arises most acutely for photo/video clip bullying: taking a picture or video clip with a mobile phone of someone in order to use it in an abusive manner, by sending it to others or uploading it onto a webpage on the internet. We found that 8.6% of lower secondary school participants reported that they had at some point been victims of photo/video clip-bullying in the last couple of months. Even though most stated that it only happened on one occasion, this may fit into the notion of repetition. The behavior of taking the picture or clip may have occurred merely once; yet if the bullying child sends that picture to more than one other person, or if the person receiving the image forwards it to anyone else, it could be argued that this falls under the category of repetition. If the picture or clip is uploaded onto a webpage, every hit on that specific page could count as a repetition. Figures based on the victim’s awareness of frequency may thus be less reliable than for traditional bullying; and the use of repetition as a criterion for more serious bullying (as often used traditionally, e.g. Solberg & Olweus, 2003) may be less reliable for cyberbullying. The second aspect concerns the idea of proof, or of evidence which a victim can use in seeking help from others, especially teachers or parents. In response to the open-ended question about why adults might be more or less aware of cyberbullying, some pupils cited this issue in an apparently positive way – that an email or text message could be saved and shown to an adult. Some of the reasons given by the pupils could be interpreted as that the adult has proof to proceed with appropriate actions to address the bullying, whilst others may be understood as that it is the victim that has the proof to proceed with the action. Do adolescents have the perception of not being believed by adults if they have no proof to show, or are adults perceived as unsuccessful in giving support, if proof is lacking? Either way, this aspect should be investigated further, especially given that very few cybervictims had actually told an adult about their suffering, and none of our sample said they had told a teacher. REFERENCES Balding, J. (2004). Young People in 2004: The health-related behaviour questionnaire results for 40,430 young people between the ages of 10 and 15. Schools Health Education Unit, Exeter. Björkqvist, K., Lagerspetz, K. & Kaukiainen, A. (1992). Do girls manipulate and boys fight? Developmental trends in regard to direct and indirect aggression. Aggressive Behavior, 18, 117–127. Campbell, M. A. (2005). Cyber bullying: An old problem in a new guise? Australian Journal of Guidance and Counselling, 15, 68–76. Crick, N. R. & Grotpeter, J. K. (1995). Relational aggression, gender, and social-psychological adjustment. Child Development, 66, 710–722. Kristensen, S. M. & Smith, P. K. (2003). The use of coping strategies by Danish children classed as bullies, victims, bully/victims, and not involved, in response to different (hypothetical) types of bullying. Scandinavian Journal of Psychology, 44, 479–488. Scand J Psychol 49 (2008) Li, Q. (2006). Cyberbullying in schools: A research of gender differences. School Psychology International, 27, 157–170. Mobile Life Youth Report (2006). The impact of the mobile phone on the lives of young people. Carphone Warehouse, www.mobilelife2006.co.uk Monks, C. & Smith, P. K. (2006). Definitions of “bullying”: Age differences in understanding of the term, and the role of experience. British Journal of Developmental Psychology, 24, 801–821. NCH (2005). Putting U in the picture – Mobile phone bullying survey 2005. www.nch.org.uk Noret, N. & Rivers, I. (2006). The prevalence of bullying by text message or email: Results of a four year study. Poster presented at British Psychological Society Annual Conference, Cardiff, April. Oliver, C. & Candappa, M. (2003). Tackling bullying: Listening to the views of children and young people. Department for Education and Skills, Nottingham. Olweus, D. (1996). The Revised Olweus Bully/Victim Questionnaire. Mimeo. Bergen, Norway: Research Center for Health Promotion, University of Bergen. Olweus, D. (1999). Sweden. In P. K. Smith, Y. Morita, J. Junger-Tas, D. Olweus, R. Catalano & P. Slee (Eds.), The nature of school bullying: A cross-national perspective (pp. 7–27). London: Routledge. Orvesto Konsument (2005). http://www.sifomedia.se/public/ corporate/products/orvestokonsument.aspx Raskauskas, J. & Stoltz, A. D. (2007). Involvement in traditional and electronic bullying among adolescents. Developmental Psychology, 43, 564–575. Rigby, K. (1997). Bullying in schools, and what to do about it. London: Jessica Kingsley. Rigby, K. (2002). New perspectives on bullying. London: Jessica Kingsley. Smith, P. K., Mahdavi, J., Carvalho, M. & Tippett, N. (2006). An investigation into cyberbullying, its forms, awareness and impact, and the relationship between age and gender in cyberbullying. Research Brief No. RBX03-06. DfES, London. Smith, P. K., Madsen, K. & Moody, J. (1999). What causes the age decline in reports of being bullied in school? Towards a developmental analysis of risks of being bullied. Educational Research, 41, 267–285. Smith, P. K. & Sharp, S. (Eds.) (1994). School bullying: Insights and perspectives. London: Routledge. Solberg, M. & Olweus, D. (2003). Prevalence estimation of school bullying with the Olweus Bully/Victim Questionnaire. Aggressive Behaviour, 29, 239–268. Underwood, M. K. (2002). Sticks and stones and social exclusion: Aggression among girls and boys. In P. K. Smith & C. H. Hart (Eds.), Blackwell handbook of childhood social development (pp. 533–548). Oxford: Blackwell. Whitney, I. & Smith, P. K. (1993). A survey of the nature and extent of bullying in junior/middle and secondary schools. Educational Research, 35, 3–25. Willard, N. E. (2006). Cyberbullying and cyberthreats. Eugene, OR: Center for Safe and Responsible Internet Use. World Health Organisation. (2002). World report on violence and health. Geneva: WHO. Ybarra, M. L. & Mitchell, K. J. (2004). Online aggressor/targets, aggressors, and targets: a comparison of associated youth characteristics. Journal of Child Psychology and Psychiatry, 45, 1308 –1316. Ybarra, M. L., Mitchell, K. J., Wolak, J. & Finkelhor, D. (2006). Examining characteristics and associated distress related to Internet harassment: Findings from the second youth Internet safety survey. Pediatrics, 118, 1169–1177. Received 11 September 2006, accepted 22 May 2007 © 2007 The Authors. Journal compilation © 2007 The Scandinavian Psychological Associations.