Methodological issues in school surveys of adolescent dating violence

Background: Research indicates that poor academic performance is associated with dating violence and with exposure to violence in the home and community. Little is known about methods in high school surveys of dating violence. An opportunity arose during a survey of dating violence prevalence in high schools in New York City to explore the relationship between academic performance, attendance and response rates and whether voice gender in Audio Computer-Assisted Self-Interview (ACASI) influenced disclosure rates.

Methods: In one school, 69% of students participated and 26% did not respond. 5% actively refused. Class failure was used as a proxy for academic performance to investigate whether responders were academically different than non-responders. The survey used ACASI in two schools, and respondents were randomized to either male or female voices.

Results: Participants were more likely to have passed all classes than non-responders (44.6% vs. 30.3%) and a test for trend was significant. The mean number of courses failed for non-responders (2.61) was significantly higher than for participants (1.59). Excellent attendance was also significantly associated with participation. The prevalence of disclosure of current and lifetime dating violence did not vary by ACASI voice gender for either boys or girls.

Conclusions: Participants are more likely to perform well in school. Since dating violence is associated with academic performance, researchers need to take account of this in estimating the degree of dating violence experienced. In contrast, students do not seem to be affected by the gender of voices in ACASI surveys of dating violence.