Music video viewing characteristics assessed included the primary type of rap music videos viewed (gangsta, bass, or hip-hop), with whom adolescents usually viewed rap music videos, and where the rap music videos were viewed.
Covariates assessed included age, employment status, involvement in extracurricular activities, participation in religious events, family composition, family’s receipt of public assistance, parental monitoring of adolescents’ whereabouts,5 and group assignment to either the HIV intervention or the comparison condition.
Outcomes
Health risk behaviors assessed whether adolescents had hit a teacher, been involved in a fight, been arrested, used alcohol or drugs (either tranquilizers, marijuana, amphetamines, lysergic acid diethylamide [LSD], cocaine, or crack), had multiple sex partners, or used condoms. Adolescents were also tested for 3 sexually transmitted diseases (chlamydia, trichomoniasis, and gonorrhea).6–8
Data Analysis
Univariate analyses described music video viewing characteristics at baseline. Subsequent bivariate analyses examined the relations among adolescents’ level of exposure to rap music videos at baseline, potential covariates, and the occurrence of health risk behaviors over the 12-month follow-up. Health risk behaviors and covariates significantly associated (P < .05) with exposure to rap music videos in bivariate analyses were included in logistic regression analyses. A separate logistic regression analysis was conducted to examine the relation between level of exposure to rap music videos at baseline and the occurrence of each health risk behavior over the 12-month follow-up. All logistic regression analyses controlled for covariates and the