Methods: We collected data as part of Healthy Teens, a longitudinal, mixed-methods study that followed a cohort of students yearly from Grades 6 to 12. Students attended one of nine middle schools, which fed into eight high schools in Northeast Georgia. We report results from surveys and focus groups when the students were in the ninth grade because it was the first year we collected quantitative and qualitative data. The final sample (surveys) comprised 624 students, 90 of whom participated in 12 focus groups (randomly sampled within gender groups). Students completed the surveys online in the schools' media centers and received a US$20 gift card. We conducted focus groups on site and gave students a healthy snack. The university's Institutional Review Board approved all protocols. For the regression analysis of the survey data we used a generalized mixed model with school as a random effect; statistical analyses were performed with SAS 9.2 PROC GLIMMIX, and we analyzed the focus group data using the constant comparative method.
Results: Findings revealed more support for female-to-male aggression, greater acceptance of norms supporting dating violence by non-Caucasian students, a strong association between norms and physical aggression but only in males, and a high correlation between victimization and perpetration. Participants rejected male-to-female dating aggression because of peer pressure not to hit girls, parents' beliefs that denounce dating violence, the superior physical advantage of boys over girls, and legal consequences.
Conclusions and Implications: The mutually supportive findings in this mixed-methods study point to the need to examine the content of evidence-based dating violence programs. To do justice to a phenomenon as complex as dating violence these curriculums must be reviewed regularly and infused with incremental gains in knowledge about cultural, social, and gender-based influences on behavior. Our findings also underscore a critical need for culturally-sensitive prevention programs developed for specific populations, differentiated by gender, race/ethnicity, and other cultural and economic influences.