This study evaluated a community-based intervention to help at-risk teens develop healthy, nonabusive relationships with dating partners. Participants were 158 14-16-year-olds with histories of child maltreatment who were randomly assigned to a preventive intervention group or a no-treatment control group. They completed measures of abuse and victimization with dating partners, emotional distress, and healthy relationship skills at bimonthly intervals when dating someone. Intervention consisted of education about healthy and abusive relationships, conflict resolution and communication skills, and social action activities. Growth curve analyses showed that intervention was effective in reducing incidents of physical and emotional abuse and symptoms of emotional distress over-time. Findings support involvement of youths in reducing the cycle of violence as they initiate dating in midadolescence.

Four studies examined the Conflict in Adolescent Dating Relationships Inventory (CADRI), a measure of abusive behavior among adolescent dating partners. Exploratory factor analysis was used to refine items based on high school participants with dating experience (N = 393; 49% female). Confirmatory factor analysis was used to derive and cross-validate the factor structure with participants from 10 high schools (N = 1,019, 55% female; ages 14-16). The model structure fit for all grades and both sexes, with physical abuse, verbal abuse, and threatening behavior most representative of the underlying "abuse" factor. In Studies 3 and 4, the second-order abuse factor showed acceptable test-retest reliability, partner agreement, and correlation (significant for males only) between observer ratings of dating partners' interactions and youths' CADRI scores. Results support the CADRI as a measure of abusive behavior in adolescent dating relationships.

Three mediators of the relationship between childhood maltreatment and dating violence perpetration during midadolescence (i.e., trauma-related symptoms, attitudes justifying dating violence, and empathy and self-efficacy in dating relationships) were tested over 1 year with a sample of students from 10 high schools (N = 1,317). Trauma-related symptoms had a significant cross-time effect on predicting incidents of dating violence for both boys and girls. Attitudes and empathy and self-efficacy did not predict dating violence over time, although they were correlated with such behavior at both time points. Child maltreatment is a distal risk factor for adolescent dating violence, and trauma-related symptoms act as a significant mediator of this relationship. The importance of longitudinal methodology that separates correlates from predictors is discussed.