ثبات در قربانی شدن مزاحمت سایبری نوجوانان: بررسی شیوع و ارتباط با وضعیت زورگویی - قربانی و سازگاری روانی- اجتماعی

Abstract The aims of this study were as follows: (a) to examine the possible presence of an identifiable group of stable victims of cyberbullying; (b) to analyze whether the stability of cybervictimization is associated with the perpetration of cyberbullying and bully–victim status (i.e., being only a bully, only a victim, or being both a bully and a victim); and (c) to test whether stable victims report a greater number of psychosocial problems compared to non-stable victims and uninvolved peers. A sample of 680 Spanish adolescents (410 girls) completed self-report measures on cyberbullying perpetration and victimization, depressive symptoms, and problematic alcohol use at two time points that were separated by one year. The results of cluster analyses suggested the existence of four distinct victimization profiles: “Stable-Victims,” who reported victimization at both Time 1 and Time 2 (5.8% of the sample), “Time 1-Victims,” and “Time 2-Victims,” who presented victimization only at one time (14.5% and 17.6%, respectively), and “Non-Victims,” who presented minimal victimization at both times (61.9% of the sample). Stable victims were more likely to fall into the “bully–victim” category and presented more cyberbullying perpetration than the rest of the groups. Overall, the Stable Victims group displayed higher scores of depressive symptoms and problematic alcohol use over time than the other groups, whereas the Non-Victims displayed the lowest of these scores. These findings have major implications for prevention and intervention efforts aimed at reducing cyberbullying and its consequences.

. Introduction Bullying can be defined as any unwanted aggressive behavior by another peer or group of peers, which involves an imbalance of power and is repeated or is highly likely to be repeated (Gladden, Vivolo-Kantor, Hamburger, & Lumpkin, 2014). Bullying may inflict harm or distress on a victim through physical, verbal, or social aggression. Similarly, cyberbullying refers to repetitive aggression carried out via electronic media (i.e., cell phones, Internet). More specifically, cyberbullying victimization includes, among a wide range of experiences, receiving threatening or insulting messages, e-mails or images, uploading images or disseminating rumors that are cruel or harmful to a victim’s reputation, the infiltration of someone’s online account in order to send messages that cause trouble for or endanger the victim, and “happy slapping,” in which recorded images of a person who is attacked or humiliated are disseminated (Smith, 2012 and Smith et al., 2006). Although traditional bullying and cyberbullying share several features in common, the latter differs in its anonymity, possibility of occurrence at any time of the day, and potentially larger audience (e.g., Kowalski, Morgan, & Limber, 2012). Recent research has shown that cyberbullying has deleterious consequences for its victims, such as depression, anxiety, drug and alcohol abuse, and suicide ideation and attempts (Gini and Espelage, 2014, Gámez-Guadix et al., 2013, Vieno et al., 2011 and Vieno et al., 2014; for a recent review and meta-analysis, see Kowalski, Giumetti, Schroeder, & Lattanner, 2014). For example, a recent cross-sectional study of data from approximately 24,000 young adolescent participants in the International Health Behavior in School-aged Children Survey (Vieno et al., 2014) utilized multilevel models of logistic regression (controlling for traditional bullying victimization, computer use, and demographics) to investigate the association between cybervictimization and psychological and somatic symptoms. The main results of the study showed that students who reported themselves as victims of cyberbullying were nearly twice as likely to experience psychological and somatic symptoms compared to their non-victimized peers, with this effect increasing substantially from occasional to frequent bullying. Such findings, based on a large representative sample of adolescents, have confirmed that cybervictimization is a significant risk factor for students who are frequently harassed online. However, this study was limited by its cross-sectional nature, which prevents establishing temporal relationships between variables. Although research on cyberbullying and cybervictimization has gained pace quite rapidly in recent years, to date, little is known about the specific victimization parameters and characteristics that impact victim outcomes. A key gap in the literature on cyberbullying is its current lack of studies on the stability of cybervictimization over time and the consequences that “stable” victimization can have for adolescents (Rueger et al., 2011 and Underwood and Card, 2012). Stability here refers to the repetition and consistency of incidents of victimization for a victim over a given period of time (e.g., one year) (Nylund et al., 2007 and Rueger et al., 2011). The stability of victimization may be closely associated with the perpetration of cyberbullying and bully–victim status. Bully–victim status refers to the categorization of adolescents as pure bullies (not victims), pure victims (not bullies), or bully–victims, that is, being simultaneously a bully and a victim (Haynie et al., 2001 and Olweus, 1993). Research has indeed found that being a perpetrator of cyberbullying significantly increases the likelihood of becoming a victim of cyberbullying (Kowalski et al., 2014). To begin filling the gap in the current lack of knowledge on the stability of cybervictimization and its relation to cyberbullying perpetration, the current study aims to analyze the stability of cybervictimization in a sample of adolescents over a period of one year. Moreover, we tested the association of cybervictimization with different bully–victim roles, as well as with negative outcomes, such as depression and problematic alcohol use.

3. Results 3.1. Analyses of the stability of cybervictimization First, we examined the stability of victimization over time. In this regard, the test–retest correlation, which is indicative of relative stability, was .40, p < .001. The paired t-tests of the mean levels of victimization at both measurement times were non-significant, indicating that victimization was stable. As the first objective of this study was to empirically determine the possible presence of a stable group of cyberbullying victims, we performed a cluster analysis with the variables “victimization at T1” and “victimization at T2” as grouping variables. Cluster analysis is an empirical method used to identify homogenous subgroups of individuals based on certain characteristics ( Everitt, Landau, & Leese, 2001). Individuals are grouped in clusters in order to minimize the variability of each group and to maximize the differences between groups in terms of the grouping variables. Thus, cluster analysis creates an empirical typology through which, unlike a theoretical typology, the data determine the patterns that form the groups ( Velicer, Redding, Anatchkova, Fava, & Prochaska, 2007). Using Ward’s method, we first performed a hierarchical cluster analysis to determine the number of clusters, which was then followed by a k-means cluster analysis. Ward’s clustering method ( Ward, 1963), which uses the squared Euclidean distance between each clustering variable, was used to identify the initial cluster solution. Ward’s method was especially suitable in this study, as it minimizes the within-cluster differences and maximizes between-group variance. The hierarchical cluster dendrograms (i.e., tree diagrams used to show the arrangement of the clusters) were then examined to identify a suitable number of clusters to retain. We also examined the theoretical meaningfulness of different solutions. The results indicated that a four-cluster solution was appropriate and meaningful for the data. The next step was to conduct a k-means cluster analysis to group the participants into a determined number of clusters using the previously identified four-cluster solution. The means and standard deviations of the victimization scores at Times 1 and 2 for each cluster are presented in Table 1. Cluster one, identified as the “Non-Victims” group, showed the lowest mean scores both at T1 victimization and T2 victimization. This group included 421 students (61.91% of the sample). Cluster two was defined as the “T1-Victims” group. This group presented repeated victimization at T1, but not at T2. This group included 99 adolescents (14.56% of the sample). Cluster three included 120 students (17.65% of the sample) and was labeled the “T2-Victims” group, as it presented repeated victimization at T2, but not at T1. Finally, cluster four was defined as the “Stable Victims” group because the participants of this group presented repeated experiences of victimization at both T1 and T2, with increasing frequency of victimization between the two times (t = −6.94, p < .001). This group included 40 adolescents (5.88% of the sample).