A Longitudinal Social Network Analysis of Tobacco Use and Friendship Dynamics Among First-Year College Students

• Activity Number: 279
• Type: Invited
• Date/Time: Tuesday, August 2, 2011 : 8:30 AM to 10:20 AM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Stephanie R. Land*+ and Ju-Sung Lee and Kristina Cooper
• Companies: University of Pittsburgh and Carnegie Mellon University and University of Pittsburgh
• Keywords: tobacco ; social networks ; longitudinal data

Abstract

Recent innovations in the modeling of social networks have yielded insights into the development of leading causes of mortality, such as tobacco use and obesity. This methodology has the potential to address important questions regarding the development of social networks, and the propagation of behaviors within social networks. Our study focuses on tobacco smoking with cigarettes and hookahs in the context of social networks among first-year college students. Participants each named up to 10 social contacts, many of whom were also participants, and provided tobacco use and social network data for themselves and their contacts in longitudinal assessments. The social network is densely connected, not limited to ego networks. We will present social network graphs and the longitudinal associations of tobacco behaviors with friendship dynamics, network popularity, betweenness, and eigenvector centrality. Ours is one of the largest longitudinal social network datasets available, providing rich opportunities for method development.