Improving Our Ability to Evaluate Underlying Mechanisms of Event Occurrence

This presentation introduces a statistical model that researchers can use to evaluate underlying mechanisms of change for outcomes measured in discrete-time epochs or gathered as event history data. Combining advantages from the basic mediation and discrete time survival analysis models, we propose a discrete-time survival mediation (DTSM) model that facilitates research involving mediators of event occurrence.

DTSM has the potential to help strengthen prevention research in several ways. First, estimating mediation effects with discrete-time outcomes can improve our understanding of discretized periods of risk and facilitate targeting intervention work. Second, operationalizing event occurrence as a discrete-time survival outcome presents advantages over other methodologies including: (a) the logistic and financial feasibility of periodic rather than continuous data collection, (b) the ability to evaluate both time-varying and time-invariant predictors, (c) the capacity to provide a pattern of probabilities of event participation over time and (d) the capability to effectively deal with censored observations.

Although DTSM has the potential to inform a broad variety of research efforts, we choose to present the need for the model in the context of early onset substance involvement in particular. Our reason for this choice is two-fold. Not only is substance abuse a critical public health problem affecting a variety of behavioral outcomes, but discrete-time survival methods are being underused in the study of substance use onset in particular given their potential to drive prevention efforts in the area. Specifically, a growing number of studies suggest that early-onset substance involvement, and alcohol use in particular, is associated with heightened risk for myriad adverse social, psychological and behavioral outcomes including later accelerated substance use itself. Research has also shown that adverse effects extend beyond these realms into brain development, suggesting that early onset substance involvement introduces neurological vulnerabilities in the formation of executive functioning and impulse control.

Given the heightened risk of problems associated with early substance involvement, targeting the timing of behavioral onset in prevention can act as a crucial element in confronting substance abuse problems. By addressing malleable determinants of early substance involvement, prevention efforts can potentially delay substance initiation and reduce risk for later adverse outcomes. DTSM can help guide empirical evaluation of the programs by permitting scientists to investigate programmatic impact on determinants of behavior and how those determinants contribute to subsequent risk for onset of substance involvement.