Link Between in-Person and Cyber Dating Violence over Time

Background: While mounting research has investigated the prevalence of dating violence (DV) among teens and emerging adults, relatively little is known about the role of cyber abuse (Zweig et al., 2013; 2014). Research examining the link between in-person and cyber DV are especially lacking, as is the longitudinal study of cyber DV. The purpose of this study is to 1) examine the prevalence of cyber DV in an ethnically diverse sample of emerging adults, and 2) examine whether in-person and cyber DV predict cyber DV over the following year.

Method: Waves 4 (N=781) and 5 (N=701) of an ongoing 6-year longitudinal study were used. At Wave 4, the sample was 58% female (mean age of 19), and self-identified as Hispanic (32%), White (30%), African American (26%), and other (12%). DV measures included physical and psychological DV perpetration and victimization from the Conflict in Adolescent Dating Relationships Inventory (Wolfe et al., 2001) at wave 4, and cyber DV perpetration and victimization (Picard, 2007; Zweig et al., 2013) at waves 4 and 5. Gender, ethnicity, and parent education were included in the path model. Multiple imputations procedure was employed for missing data. Because cyber DV victimization and perpetration were positively skewed, maximum likelihood estimator with robust standard errors (MLR) was employed in Mplus 7.11.

Results: Consistent with limited existing research, 18% of emerging adults reported past year perpetration of cyber DV and 24% reported past year victimization at baseline (19% and 23% at 1 year follow-up, respectively). Among in-person traditional DV, only physical DV victimization at wave 4 was positively related to cyber DV victimization at Wave 5, est = 0.33, SE = 0.12, p <.01. Cyber DV perpetration at wave 4 was also positively related to cyber DV victimization over time (est= 0.24, SE = 0.12, p <.05), whereas cyber DV victimization was not, est = 16, SE = 0.10, p >.05. On the other hand, cyber DV victimization was negatively associated with cyber DV perpetration (est = -0.12, SE = 0.05, p <.05) while cyber DV perpetration was positively related to cyber DV perpetration (est = 0.55, SE = 0.08, p <.001). Among demographic variables, only gender was longitudinally related to cyber DV perpetration, with females (est = 0.16, SE = 0.08, p <.05) more likely to perpetrate the following year.

Conclusion: As evidenced by the relatively high rate of cyber DV observed in this and other samples, as well as its longitudinal relationship to in-person DV, researchers and practitioners should include cyber DV in future research studies and interventions. Implications will be thoroughly discussed, including how current results can inform public policy around DV.