Which Victims Are at Greatest Risk of Becoming Bullies Themselves? a Propensity Score Stratification Analysis

Introduction:Violence and bullying in school is an increasing public health problem in Korea and globally. Many studies and prevention programs concerning bullying and victimization in school focus on either the perpetrators or victims. However, a major shortcoming of existing literature is that there is less focus on the bully-victims, who are at highest risk for psychological, behavioral, and academic difficulties, despite its small group size. To understand the characteristics of this understudied vulnerable group of youth, the current study examined the characteristics of the bully-victim group: who are victims with the greatest risk of becoming perpetrators themselves. Specifically, the current study explored whether the “typical” victim or the “non-typical” victim is most vulnerable to exhibiting bullying behaviors as well. 

Methods: The study applied a two-step propensity score stratification procedure using a nationally representative data from the Korea Children and Youth Panel Study 4^th Grade Panel (N=1,977). Waves3 (first year of middle school) and 4 (second year of middle school) contained information necessary for analysis. In step 1, past covariates (Wave3), such as gender, depression, aggression, school adjustment, parental abuse and neglect, family socioeconomic indicators, were used to predict the probability of current victimization (Wave4). Using the estimated propensity score, youth were stratified into four groups. The relationship between victimization (Wave4) and bullying (Wave4) were separately estimated for each group. In step 2, the strata were regressed on the four coefficients to identify an upward (i.e., the “typical” victim is most vulnerable to engage in perpetrating behavior) or downward (i.e., the “non-typical” victim is most vulnerable to exhibiting perpetrating behavior) trend in the relationship between victimization and bullying across strata.

Results: Findings indicate that the non-typical victim has greater odds of exhibiting perpetrating behavior (stratum 1: B=2.42, p <.0001; stratum 2: B=2.08, p=.003; stratum 3: B=1.53, p= .001; stratum 4: B=0.95, p=.06; overall trend: B=-0.50, p=.04). In other words, when youth who are LEAST likely to experience victimization—characterized as having lowest levels of internalizing and externalizing problems, living with least abusive and neglectful parents, showing highest levels of school adjustment—actually experience victimization, they are also MOST likely to engage in bullying behavior. In contrast, for the group of youth with psychosocial, family, and school characteristics that would make them most vulnerable to being the target of victimization, experience of victimization is minimally associated with co-experience of perpetration.                                           

Conclusions and Implications: The bully-victim group merits attention from researchers and practitioners, as this marginalized group is most vulnerable in terms of psychosocial outcomes, compared to “pure” bullies and victims. The current findings provide a detailed understanding of the bully-victims, by demonstrating the characteristics of victimized youth who have the greatest probability of also showing perpetrating behaviors. Study findings indicate that the non-typical victim is more likely to be involved in bullying activities than the typical victim, and hence fall into the category of bully-victims. This comprehensive information supports the utility of intervention programs against school violence that specifically target youth who are both the victims and the perpetrators of bullying.