Poster #217 - Bidirectional Longitudinal Relationships between Bullying/Cyberbullying and Victimization

• In Event: 1-061 - Poster Session 02
  In Poster Session: PS 02 Section - Social Relationships

Integrative Statement

Background: Bullying and cyberbullying are widespread among youth worldwide and have serious deleterious effects for adolescents and for societies at large (Zych, Ortega-Ruiz, & Del Rey, 2015). Preventative interventions to empower victims have shown limited success (Evans, Fraser, & Cotter, 2014). Ma (2001) suggests that bullies might be victims themselves as well as vice versa, thus, interventions targeting only bullying behavior or victimization separately might be ineffective. To better inform interventions, understanding the longitudinal bully-victim cycle seems paramount. Previous research has shown significant bidirectional relationships between bullying and victimization (Jose, Kljakovic, Scheib, & Notter, 2012; Marsh, Parada, Craven, & Finger, 2004); however, these findings were based on short-term longitudinal data, therefore, the longer-term bidirectional relationships remain largely unknown. Furthermore, whether there is a bidirectional relationship between cyberbullying and and cybervictimization is unknown. Given that cyberbullying is both unique but related at its core to traditional bullying behaviors (Zych et al., 2015), it seems key to separately test the relationships in both forms of bullying and victimization. The current study has two main research goals: First, to test the longitudinal bidirectional relationships between traditional bullying and victimization, and second, to test the bidirectional relationships between cyberbullying and victimization. Method: Data were recently collected four times as part of an ongoing longitudinal study among 569 adolescents located in a medium-sized city in the Czech Republic (58.6% female; Mage=12.42 years, SD=0.66 at Wave 1), twice annually over the course of two years. Bullying and cyberbullying perpetration and victimization were measured across all four assessments using the measures developed by Gradinger, Strohmeier, and Spiel (2010). Control variables included adolescent age, sex, family structure, and socioeconomic status (SES) measured at W1. Results: Descriptive statistics and bivariate correlations (see Table 1) were computed in a first step; two longitudinal cross-lagged path models were hypothesized (one for the traditional bullying-victimization cycle, the other for the cyberbullying-victimization cycle) and tested in SEM (see Figure 1), which also included stability paths. Correlation analysis showed a weak to moderate positive relationship between bullying and victimization across time points as well as between cyberbullying and victimization across most assessments. Results of cross-lagged model tests provided evidence of significant bidirectional relationships between traditional bullying and victimization: bullying at W1 predicted victimization at W2 (ß = .12, p< .05); victimization at W2 predicted bullying at W3 (ß = .18, p< .001); and bullying at W3 predicted victimization at W4 (ß = .30, p< .001). The relationship was also significant between cyberbullying and victimization: cyberbullying at W1 predicted victimization at W2 (ß = .18, p< .001); victimization at W2 predicted bullying at W3 (ß = .12, p< .05); and bullying at W3 predicted victimization at W4 (ß = .42, p< .001). Discussion: Findings provide evidence for bidirectional, longitudinal effects between bullying/cyberbullying and victimization measures over time. Future interventions should take into consideration the cyclical nature of the relationship between perpetration and victimization.

Authors

• Magda Javakhishvili, University of Kentucky
  Presenting Author

• Dan Liu, University of Kentucky
  Non-Presenting Author

• Marek Blatný, Academy of Sciences of the Czech Republic
  Non-Presenting Author

• Alexander T Vazsonyi, University of Kentucky
  Non-Presenting Author