youth tobacco use; (c) chewing gutkha is considered less harmful and more accessible than smoking cigarettes; (d) schools are not promoting tobacco control activities; and (e) students were enthusiastic about the role government should play in tobacco control. These findings are being used to develop a comprehensive intervention program to prevent and control tobacco use among Indian youth.

28:T49f,

Recent research suggests that a youth development framework emphasizing youth assets may be a promising intervention strategy for preventing adolescent risk behaviors. Understanding how neighborhood and community resources relate to youth assets may aid in identifying environmental strategies to complement individually oriented asset-building interventions. In this study, 1,350 randomly selected inner-city youth and their parents (paired interviews) were interviewed in person. After controlling for demographic characteristics of youth and parents using multivariate logistic regression, parental perception of neighborhood safety was associated with the nonparental adult role model asset, peer role model asset, and for African American youth, the community involvement asset. City services and neighborhood services were associated with use of time (groups/sports) and use of time (religion), respectively. Psychological sense of community was associated with community involvement for Native American youth. Findings suggest that neighborhood and community-level influences should be considered when designing youth development interventions to reduce risk behaviors.

29:T411,

Question-order effects refer to systematic differences in responses that can be attributed to the manner in which questions assessing attitudes and cognitions are asked. This article hypothesized that question-order effects in assessing the perceived importance of skin cancer would be moderated by the extent to which people are involved with the issue of skin cancer. A telephone survey (N = 325) was conducted by administering two versions of the same questionnaire, one in which importance of skin cancer was assessed without other contextual variables and another in which contextual variables were asked before assessing the importance of skin cancer. As hypothesized, when people were highly involved with the issue, question-order effects did not occur. When involvement was low, importance of skin cancer was greater in the absence of contextual questions than in their presence. Findings have implications for how health-related issues are communicated to the public and how formative research is conducted.

2a:T414,

Theories of health behavior posit that change is accomplished by modifying factors deemed as mediators. A set of mediators from several theoretical models used in sexual risk reduction programs was assessed among a sample of 522 African American female adolescents. The goal was to determine whether self-esteem was associated with sexually transmitted disease (STD), pregnancy, and the set of theoretical mediators controlling for covariates. Bivariate analyses showed no relationship between self-esteem and STD or pregnancy; multivariate regression analysis revealed a significant relation between self-esteem and the set of mediators. Girls higher in self-esteem were more likely to hold positive condom attitudes, felt more efficacious in negotiating condom use, had more frequent communication with sex partners and parents, perceived fewer barriers to using condoms, and were less fearful of negotiating condom use. Self-esteem should be considered when designing and evaluating sexual risk reduction programs for this population.

2b:T6fd,