Dana Haynie



The Dynamic Effect of Friendship Networks on Adolescent Risky Sexual Behavior


Principal Investigator:
Dana Haynie (Sociology & CJRC)

Kristie Williams (Sociology)
Brian Soller (Sociology PhD student & CJRC)
Nate Doogan (Social Work PhD student & CJRC)
Christopher Keenan (Sociology PhD student & CJRC)


Reducing risky sexual behavior (RSB) among adolescents is a national priority and understanding the social context in which adolescents make decisions about sexual behavior is critical to designing effective interventions. Although it is generally accepted that peers play a fundamental role in shaping adolescent behavior in general, the role of friendship networks in shaping RSB in particular, is poorly understood. Identifying the complex ways in which friendship networks shape sexual behavior in adolescence is essential to designing and targeting interventions to improve adolescent health and reduce health disparities. The proposed study advances research by using complete longitudinal network data from the National Longitudinal Study of Adolescent Health along with innovative, actor-oriented network models (SIENA) to examine the reciprocal influence of friendship networks on change in risky sexual behavior over time. This makes possible several innovations including the estimation of selection and influence processes regarding the impact of friends' behavior/attitudes on adolescent involvement in RSB, incorporating measures of friends' reported behaviors/attitudes rather than relying on respondent perceptions, considering the relative importance of friends' attitudes versus friends' behavior when it comes to determining influence processes, considering how the impact of friends' behavior/attitudes depend upon prosocial influences in adolescents' lives (e.g., parenting practices, religion, school involvement) and examining whether friendship networks can account for race, gender, and SES disparities in adolescents' experiences of RSB. As a result, we will be able to address the following three aims: 1) Determine how friendship network structure, attitudes, and behavior affect adolescent involvement in risky sexual behavior, 2) identify the role of friendship networks in contributing to race/ethnic, gender, and class differences in risky sexual behavior, and 3) identify which adolescents are most susceptible to the influence of friends on sexual behavior, with attention to the moderating role of parents and other prosocial influences.

The current study fills a critical need by using longitudinal data on friendship networks to examine the role of friends for predicting adolescent involvement in risky sexual behavior net of selection effects and of other parenting and prosocial factors. Results from this study will be used to design interventions aimed at reducing disparities in risky sexual behavior and improving the health of all adolescents.