Session: Preventing Risky Behavior Across Development: Applying Innovative Longitudinal Methods to Strengthen Prevention Science (Society for Prevention Research 22nd Annual Meeting)

The goal of this symposium is to present three innovative longitudinal methodologies that can be applied to enhance the understanding of the etiological processes underlying the development of risky behavior.  The symposium supports the conference themes of epidemiology/etiology and innovative methods and statistics.  The three papers represent a diversity of approaches and applications to prevention science.

The first paper, “Peer Socialization of Aggressive Behavior in Elementary School:  Using Propensity Scores to Strengthen Causal Inference” uses inverse propensity weighting (IPW; Hirano & Imbens, 2001) and social network analysis (SNA; Wasserman & Faust, 1994) to examine the effect of friends’ aggressive behavior on individual aggressive behavior in elementary school.  IPW can help prevention scientists to more accurately determine the causal effects of risk factors by mimicking randomization when randomization is not possible or not ethical (i.e., randomization to peer groups in this case). 

The second paper, “Preventing Adolescent Alcohol Use and Delinquency:  A Dynamical Systems Analysis of Genetic Moderation of Intervention Effects” uses a dynamic systems approach (Boker, 2007) to examine genetic moderation of the effect of the PROSPER intervention on substance use and delinquency in middle adolescence.  Dynamical systems modeling can help prevention scientists to evaluate the efficacy and effectiveness of prevention programs from a new perspective. Particularly, they may be used to investigate the change of the slope (i.e., acceleration) and its relation with levels and slopes, revealing potential prevention effects on the system parameters themselves. Dynamical system models can also help prevention scientists to uncover the intrinsic self-regulating system and extrinsic environmental influences (i.e., coupling with other behaviors), shedding light on how prevention programs prevent the onset and escalation of problem behavior. 

The third paper, “Using the Time-Varying Effect Model to Examine Changes in Predictors of Multiple Sexual Partners from Adolescence through Young Adulthood” demonstrates use of time-varying effect modeling (TVEM; Tan et al., 2012).  TVEM allows for the modeling of near-continuous time.  The use of TVEM can help prevention scientists to determine which risk factors are particularly salient at different ages and can point to particularly useful developmental periods for intervention. 

A discussion is designed to highlight the benefits of using innovative longitudinal methods and their contribution to theory, research, and practice related to the understanding of the etiological processes underlying the development of risky behavior.