مزاحمت سایبری در نوجوانان: روش ها و مشخصات متجاوزان
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|30346||2010||8 صفحه PDF||سفارش دهید||محاسبه نشده|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Computers in Human Behavior, Volume 26, Issue 5, September 2010, Pages 1128–1135
In this study, a questionnaire (Cyberbullying Questionnaire, CBQ) was developed to assess the prevalence of numerous modalities of cyberbullying (CB) in adolescents. The association of CB with the use of other forms of violence, exposure to violence, acceptance and rejection by peers was also examined. In the study, participants were 1431 adolescents, aged between 12 and17 years (726 girls and 682 boys). The adolescents responded to the CBQ, measures of reactive and proactive aggression, exposure to violence, justification of the use of violence, and perceived social support of peers. Sociometric measures were also used to assess the use of direct and relational aggression and the degree of acceptance and rejection by peers. The results revealed excellent psychometric properties for the CBQ. Of the adolescents, 44.1% responded affirmatively to at least one act of CB. Boys used CB to greater extent than girls. Lastly, CB was significantly associated with the use of proactive aggression, justification of violence, exposure to violence, and less perceived social support of friends.
Violent behavior among adolescents and young people is a severe problem in many countries. In recent years, new forms of aggression based on information and communication technology (computers, cell phones, etc.) have been added to the traditional forms of violence. In this context, cyberbullying (CB) has been defined as an aggressive and deliberate behavior that is frequently repeated over time, carried out by a group or an individual using electronics and aimed at a victim who cannot defend him- or her-self easily (Smith, 2006). Patchin and Hinduja (2006) describe it as deliberate and repeated harm performed with some kind of electronic text. These violent behaviors can be carried out by means of a cell phone, electronic mail, Internet chats, and online spaces such as MySpace, Facebook, and personal blogs. Although in many cases, CB implies acts of traditional aggression (for example, insulting, spreading rumors, or threatening), which are communicated electronically instead of face-to-face, CB can also include unique behaviors with no analogue in traditional bullying. For example, the phenomenon known as bombing occurs when the aggressor uses an automated program to collapse the victim’s e-mail with thousands of simultaneous messages, causing failure and blocking of the victim’s e-mail account ( Burgess-Proctor, Patchin, & Hinduja, 2008). As this phenomenon is new, there is as yet little agreement about the diverse categories of this form of violence, so that in the studies carried out, different classifications can be found (e.g., Burgess-Proctor et al., 2008, Smith et al., 2006, Willard, 2006 and Willard, 2007). For example, according to Willard, 2006 and Willard, 2007, some of the modalities that CB can adopt are (1) online fights, known as flaming, which imply the use of electronic messages with hostile and vulgar language; (2) slandering, a modality that implies online disparagement, for example, sending cruel images or rumors about others to spoil their reputations or social relationships; (3) impersonation (hacking) by infiltration into someone’s account in order to send messages that make the victim lose face, cause trouble for or endanger the victim, or harm the victim’s reputation and friendships; (4) defamation by spreading secrets or embarrassing information about someone; (5) deliberate exclusion of someone from an online group; (6) cyber harassment or the repeated sending of messages that include threats of injury or that are very intimidating. The phenomenon known as happy slapping consists of recording with cell phone cameras images in which a person, who is often in a minority situation, is attacked. The image or video is later shared with friends, posted online, or distributed electronically. This phenomenon has recently been the object of attention by the mass media in many European countries. Diverse studies warn about the high occurrence of CB. Table 1 shows a sample of some representative studies. For example, in one of the first studies of CB, Ybarra and Mitchell (2004) surveyed 1501 children and adolescents between ages 10 and 17 years by phone and found that 12% had participated in cyberbullying. Li (2006), in a sample of 264 high school students, found that 22% of the boys and 12% of the girls admitted having cyberbullied others. In a cross-cultural study, the same authoress found percentages of cyberbullies ranging between 15% and 7%, respectively, for Canada and China (Li, 2008). In a study carried out with Turkish students, 35.7% admitted performing CB (Aricak et al., 2008). In Spain, Ortega, Calmaestra, and Mora-Merchán (2008) found in adolescents that 5.7% admitted having performed CB occasionally, and 1.7% had carried out severe forms of CB. Although most of the research was carried out in schools, in some cases it was online (Hinduja & Patchin, 2008). Table 1. Representative studies of cyberbullying. Authors Measurement of CB Sample Prevalence rate Aricak et al. (2008) “Questionnaire of Cyberbullying” (QoCB) 269 students, aged between 12 and 19 years from Istanbul 35.7% Beran and Li (2005) Questionnaire of 15 items, based on a definition of harassment. They use open questions to specify types. Closed for frequency, being victim/aggressor, emotional and behavioral responses 432 students from 7th to 9th grade. Canada 22% once or twice; 4% several times. No gender differences Dehue et al. (2008) They created a questionnaire for students and one for parents. The questionnaire for students, among other things, asks about the prevalence and the methods used to carry out and/or be a victim of CB (SMS, e-mail, gossiping, ignoring, hacking, name-calling), the anonymity of the aggressor and the sex of the victim. The questionnaire is based on the based on “Olweus Bully/victim Questionnaire” and on the “Amsterdam Bullying Questionnaire for Children” 1211 participants from the last grade of primary education and the first of secondary education. Mean age: 12.7 Holland 16% significantly more boys (18.6%) in comparison with girls (13.4%) Li (2006) An anonymous survey that, among other aspects, asks whether the respondent had cyber-bullied others. If so, they should indicate the means (e-mail, chat room, cell phone, other) and the frequency 264 students from 7th to 9th grade. Canada 17% (22.3% boys and 11.6% girls) Li (2007a) The same questionnaire as in Li (2006) 177 Canadian students from 7th grade. Canada 14.5% Li (2007b) The same questionnaire as in Li (2006) 461 students from 7th grade. Canada and China 17.8% (21.9% boys and 13.4% girls) Li (2008) The same questionnaire as in Li (2006) 354 students between 11 and 15 years. Canada and China Canada: 15% China: 7% Ortega et al. (2008) Cyberbullying Questionnaire (Ortega, Calmaestra, & Mora-Merchán, 2007). Includes questions about CB through cell phones and the Internet 830 Spanish students between 12 and 18 years 1.7% severe cyberbullying, 5.7% occasionally or moderate No gender differences Hinduja and Patchin (2008) Online. First, a description of CB. Includes two questions: Have you ever performed CB against others? and Have you ever threatened to physically harm someone or have you scared others this way online? 680 boys/698 girls 10–17 years 18% boys and 15.6% girls Patchin and Hinduja (2006) Online. First, a description of CB. Includes two questions: Have you ever performed CB? and, Have you ever threatened to physically harm or scared others this way online? 384 people less than 18 years of age 10.7% Smith et al. (2008) Questionnaire based on Olweus’ Bully/Victim Questionnaire (Solberg & Olweus, 2003). The questionnaire includes a definition of bullying followed by a statement about cyberbullying as including the seven media: through text messaging, pictures/photos or video clips, phone calls, e-mail, chat rooms, instant messaging, and websites. Next, it asks about the frequency of CB, for each of the seven media. Open-ended questions allowed pupils to give more detailed answers on examples of cyberbullying, reasons for perceived impact, and suggestions for stopping it. The time-frame was the “past couple of months” Two surveys with pupils aged 11–16 years: (1) 92 pupils, supplemented by focus groups; (2) 533 pupils 6.6% often (2 or 3 times a month, once a week, or several times a week) and 15.6% once or twice No gender differences Ybarra and Mitchell (2004) “Youth Internet Safety Survey”. By telephone. Included two questions for aggressors: (1) Have you ever made disagreeable or vulgar remarks about others on the Internet? (2) Have you used the Internet to harass or shame someone with whom you were angry? 1501 young people between 10 and 17 years USA 12% Williams and Guerra (2007) They use one item: I told lies about some students though e-mail or instant messaging 3339 youths in 5th , 8th, and 11th grades 9.4% Table options A limitation of many of the studies is that they assessed the occurrence of CB generically, without specifying in detail the modalities employed. For example, in some cases, they focused on providing a definition of CB and asking the participants whether they had carried out CB, and if so, to describe aspects such as the means employed (chat room, e-mail, cell phone). This perspective is valuable to address a new phenomenon about which relatively few studies have been carried out. However, as we discover more about the importance of the phenomenon, it becomes appropriate to develop more specific measures that include a broad array of CB modalities. Therefore, the first goal of this study consisted of developing a questionnaire to assess the performance of many types of CB by adolescents. On the other hand, CB reveals a series of differences with the traditional types of maltreatment and bullying among schoolmates. Elements such as the perception of online anonymity and the safety of hiding behind a computer screen contribute to freeing individuals from traditional constraints and social pressures, as well as from moral and ethical misgivings (Hinduja and Patchin, 2008, Li, 2007a and Li, 2007b). Thus, adolescents who would not behave violently in a face-to-face situation can adopt different roles and perform this type of violence. Anonymity also implies the absence of consequences, because the aggressors frequently cannot be identified and, therefore, they avoid punishment. These characteristics make one wonder whether the psychological profile of the adolescents who carry out CB is similar to or different from the profile associated with traditional forms of violence. Therefore, the second goal of the study consisted of assessing the relation between CB, offline forms of violence, and associated risk factors. Regarding the relation of CB to other violent behaviors in adolescents, CB should be associated with forms of proactive aggression and indirect aggression. Proactive aggression consists of deliberate and planned behavior with the intention of obtaining a reward and is differentiated from reactive aggression, which refers to a furious response to a perceived threat or provocation (Dodge, 1991). In fact, previous studies suggest that traditional bullying is more closely associated with proactive than with reactive aggressiveness (Roland and Idsøe, 2001, Schwartz et al., 1998 and Unnever, 2005). On the other hand, indirect aggression, also called relational or social aggression, consists of harming someone by means of manipulating relationships (Björkqvist, 2001, Björkqvist et al., 1992 and Crick and Grotpeter, 1995). In this type of aggression, covert strategies are used in order to exclude and isolate rivals in the peer group. These actions include spreading rumors about others, threatening to end personal relationships, and to reveal private information (Crick, 1995 and Galen and Underwood, 1997). In this sense, certain forms of CB have the same characteristics as traditional indirect bullying (Dehue, Bolman, & Völlink, 2008) and CB has even been defined as a computer-mediated form of indirect aggression (Piazza & Bering, 2009). One of the risk factors that have been traditionally associated with the above-mentioned forms of aggressive behavior, and with bullying in particular, are normative beliefs about the justification of violence. Adolescents’ experiences throughout their lives lead them to store in their memories certain knowledge structures that affect their future behavior. Cognitive-social theories have generally called such knowledge structures schemas or scripts (Huesmann, 1988). In the case of aggressive behavior, many studies have revealed the presence of schemas related to the justification of the use of violence. For example, various studies have detected that children and adolescents who believe that it is appropriate to attack others when they deserve it are more apt to be aggressive (Bentley and Li, 1995, Bosworth et al., 1999, Calvete, 2008, Calvete and Cardeñoso, 2005 and Huesmann and Guerra, 1997). In fact, a recent study of Williams and Guerra (2007) found an association between justification of violence and CB. However, in their study, the measurement of CB was a single question that referred to spreading lies about classmates by e-mail or instant messaging, and other forms of CB were not included. In addition, a series of contextual variables have been linked to violent behavior for a long time. Firstly, the role of exposure to violence was pointed out from the social learning model by Bandura (1986). Children who observe more positive consequences and fewer negative ones for aggression acquire the belief that aggressive behavior leads to good consequences. In general, diverse studies support the fact that aggressive behavior increases with exposure to violence at home, at school, in the neighborhood, and in the mass media (Baldry, 2003, Coyne and Archer, 2005, Flannery et al., 2004, Gorman-Smith et al., 2004, Harold and Conger, 1997, Huesmann et al., 2003 and Schwartz and Proctor, 2000). Lastly, among the contextual variables are experiences with peers, and rejection by peers is one of the most important factors (Laird, Jordan, Dodge, Pettit, & Bates, 2001). Numerous investigations have found a clear relation between aggressive behavior and rejection (see Dodge, Coie, & Lynam, 2006). However, the results have been mixed, depending on the type of aggressive behavior (Card and Little, 2006, Price and Dodge, 1989 and Salmivalli et al., 2000), and, in general, suggest that rejection is positively associated with reactive aggression and negatively with indirect aggression. Summing up, the goals of this study were (1) to develop a questionnaire to assess a variety of CB behaviors. This goal in turn implied the assessment of the measurement model of the instrument. (2) To study the relation of CB with other indicators of aggressive behavior such as the justification of violence and the frequency of proactive, reactive, direct, and indirect aggressive behaviors. And (3) to study the association between CB and diverse contextual variables such as exposure to violence and acceptance and rejection by peers.
نتیجه گیری انگلیسی
3. Results 3.1. Factor analysis of the Cyberbullying Questionnaire The Kaise–Meyer–Olkin index was .96, indicating that the correlation matrix was suitable for factor analysis. The parameters for confirmatory factor analysis were estimated using the polychoric and the asymptotic covariance matrixes of the CBQ items. We tested a one-factor model via Weight Least Squared estimation with LISREL 8.8 (Jöreskog & Sörbom, 2006). Following the recommendations from a number of authors (e.g., Hu & Bentler, 1999), goodness of fit was assessed by the comparative fit index (CFI), the nonnormed fit index (NNFI), and the root mean square error of approximation (RMSEA). Generally, CFI and NNFI values of .95 or above and RMSEA values less than .06 reflect good fit. In all the models, we used the effects-coding method proposed by Little, Slegers, and Card (2006) to identify and set the scale of the latent variables. This method consists of constraining the set of indicator intercepts to sum to zero for each construct and the set of loadings for a given construct to average 1.0, which is the same as having them sum to the number of unique indicators. According to Little et al., this method is suitable to confirm the factor structure of a construct from particular items. The fit indexes were excellent for the model, χ2(104, n = 1431) = 140, RMSEA = .016 (0.0079; 0.022), NNFI = 1, CFI = 1. All the factor loadings ranged between .90 and .99. Alpha coefficient was .96 and the mean correlation between items was .64, revealing that, in general, the use of the diverse CB modalities was highly consistent. 3.2. Frequency of CB behaviors Of the sample, 44.1% of the adolescents responded affirmatively to at least one of the items of the CBQ. Table 2 presents the prevalence rates of each type of CB. The most frequent behaviors were intentionally excluding a classmate from an online group (20.2%), hanging jokes, rumors, gossip, or embarrassing comments about a classmate on the Internet (20.1%), sending the link of such comments to others (16.8%), and hacking in order to send messages by e-mail that could cause trouble for the victim (18.1%). The two forms of CB known as happy slapping (filming someone while they are forced to do something humiliating or filming someone while they are being attacked) were indicated by 10.4 and 10.5%, respectively, of the adolescents. Table 2. Prevalence rates per items. Sometimes Often Total 1. Sending threatening or insulting messages by e-mail 14.2 1.6 15.8 2. Sending threatening or insulting messages by cell phone 14.2 1.5 15.7 3. Hanging humiliating images of classmates on the Internet 9.0 1.0 10 4. Sending links of humiliating images to other people for them to see 8.2 0.9 9.1 5. Writing embarrassing jokes, rumors, gossip, or comments about a classmate on the Internet 18.3 1.8 20.1 6. Sending links with rumors, gossip, etc. about a classmate to other people so they can read them 15.5 1.3 16.8 7. Hacking to send messages by e-mail that could make trouble for the other person 15.5 2.6 18.1 8. Recording a video or taking pictures by cell phone while a group laughs and forces another person to do something humiliating or ridiculous 9.4 1.0 10.4 9. Sending these images to other people 9.6 1.5 11.1 10. Recording a video or taking pictures by cell phone while someone hits or hurts another person 9.6 0.9 10.5 11. Sending these recorded images to other people 10.0 1.3 11.3 12. Broadcasting online other people’s secrets, compromising information or images 13.3 1.3 14.6 13. Deliberately excluding someone from an online group 18.1 2.1 20.2 14. Sending messages massively that include threats or are very intimidating 8.3 0.9 9.2 15. Recording a video or taking cell phone pictures of classmates performing some kind of behavior of a sexual nature 8.4 0.7 9.1 16. Sending these images to other people 7.7 0.9 8.6 Table options The content of the responses to the open questions were analyzed. Examples of Items 3 and 4 were “changing clothes in the locker room”, “bathing nude in the river”, “manipulated photographs, adding, for example, a moustache”, or “silly dancing”. Examples of Items 8 and 9 were “cutting off the leg of a chair so they will fall down when they sit and then recording them” and “making someone sing something silly and sending the video”. Most of the contents of Items 10 and 11 refer to kicking classmates, or, in some cases, a vagabond. Lastly, in some cases, the adolescents indicated that they recorded aggressive scenes but as a joke, like a stage scene. In these cases, it was not considered CB and these responses were not included in the prevalence rate or in the other statistical analyses. 3.3. Variables associated with CB behavior Gender differences were found in the use of CB. With regard to the prevalence rate, 40.3% of the girls and 47.8% of the boys responded affirmatively to at least one of the CBQ items, and the difference was statistically significant, χ2(1, n = 1403) = 7.95, p < .01. When assessing the differences by item, we found that they occurred in the behaviors of recording humiliating images of classmates, χ2(2, n = 1380) = 8.45, p < .05, recording physical aggressions, χ2(2, n = 1380) = 10.41, p < .01, sending images of physical aggression, χ2(2, n = 1380) = 8.72, p < .05, and sending images of a sexual nature, χ2(2, n = 1380) = 8.27, p < .05. The differences were observed especially for the response category “often”. So, 1.8% of the boys versus 0.1% of the girls often recorded physical aggressions, 2% of the boys versus 0.4% of the girls recorded humiliating images of a classmate, 2.14% of the boys versus 0.4% girls sent images of the recorded physical aggressions, and 1.5% of the boys versus 0.1% of the girls sent images of classmates of a sexual nature. The total frequency of CB was calculated by adding the responses to the 16 CBQ items and an analysis of variance was carried out to assess the differences as a function of gender and school level. Table 3 shows the means and standard deviations obtained. For school grade, statistically significant differences were observed, F(3, 1390) = 25, p < .001. Subsequent comparisons (Tukey’s method) indicated that the frequency of CB behaviors was significantly higher in second and third grade of secondary education than in first and fourth grade, p < .05. There were no significant differences between second and third grade. No significant gender differences were found in CB frequency, M = 2.15 and SD = 5 in girls versus M = 2.53 and SD = 4.99 in boys, F(1, 1390) = 2.69, ns. However, boys scored higher than girls in all the other aggressive behaviors (proactive, reactive, direct and indirect). Table 3. Frequency of CB as a function of academic grade and gender. First grade of ESO n = 389 Second of ESO n = 353 Third grade of ESO n = 369 Fourth grade of ESO n = 300 Total Female 0.76 (1.81) 4.17 (6.98) 2.40 (5.08) 1.19 (1.75) 2.15 (4.73) Male 1.27 (2.69) 3.41 (5.63) 3.46 (6.28) 2.05 (4.38) 2.53 (5.00) Mean (SD) 1.02 (2.32) 3.82 (6.39) 2.93 (5.73) 1.56 (3.19) 2.33 (4.86) Note: ESO, Educación Secundaria Obligatoria (compulsory secondary education). Table options To assess the association between CB behavior and other indicators of aggressive behavior, we carried out multiple regression analysis including the scores in proactive aggressive behavior, reactive aggressive behavior, direct aggressive behavior, indirect/relational aggressive behavior, and justification of violence as predictor variables. This model explained 13% of the variance (R2 = .13, p < .001). Table 4 shows the results. The total CB score was only significantly associated with proactive aggressiveness and beliefs that justify violence. Table 4. Indicators of aggressive behavior associated with CB. B SE β t Proactive aggressiveness .52 .06 .34 8.29⁎⁎ Reactive aggressiveness −.05 .06 −.04 −0.90 Direct aggressiveness −.06 .06 −.03 −0.93 Relational aggressiveness −.00 .07 .00. −0.06 Justification of violence .07 .03 .08 2.20⁎ ⁎ p < .05. ⁎⁎ p < .001. Table options Lastly, another regression model was estimated, including as predictor variables the contextual variables rejection by others, acceptance by others, perceived social support, and exposure to violence. This model only explained 2.5% of the variance of CB (R2 = .025, p < .001). Table 5 shows the results. CB was positively associated with exposure to violence and negatively with perceived social support. Table 5. Contextual variables associated with CB. B SE β t Rejection by others .003 .17 .00 0.02 Acceptance by others .031 .17 .01 0.19 Perceived social support −.122 .04 −.10 −3.10⁎ Exposure to violence .057 .01 .13 4.21⁎⁎ ⁎ p < .05. ⁎⁎ p < .001.