December 19, 2011
Linda M. Collins, PhD
Director of The Methodology Center and Professor of Human Development and Family Studies, Pennsylvania State University
About the Presentation: Behavioral interventions are used widely for prevention and treatment of health problems and for promotion of health. Behavioral interventions are typically developed and evaluated using a treatment package approach. In this approach the intervention is assembled a priori and evaluated by means of a two-group randomized controlled trial (RCT). Refinement of an intervention is often done by conducting post-hoc analyses on data from the RCT. In this talk I suggest an alternative framework for building and evaluating behavioral interventions. This new framework, called the Multiphase Optimization Strategy (MOST), is a principled approach to intervention optimization that has been inspired by ideas from engineering. MOST includes the RCT for intervention evaluation, but also includes other steps before the RCT aimed at intervention optimization. Using MOST, behavioral interventions can be optimized using criteria chosen by the intervention scientist. The goal may be to develop a cost-effective intervention; an intervention that achieves a specified level of effectiveness; the briefest intervention that achieves a minimum level of effectiveness; or any other reasonable goal. The MOST framework relies heavily on resource management by strategic choice of highly efficient experimental designs. I propose that MOST offers several benefits, including more rapid long-run improvement of interventions, without requiring a dramatic increase in intervention research resources.
About the Presenter: Dr. Collins’ current interests include phased experimental approaches for optimization of behavioral interventions for prevention and treatment of health disorders; applying ideas from engineering, such as control theory, to intervention optimization; and statistical methods for longitudinal research, particularly Latent Transition Analysis (LTA), a method for fitting models of discrete development.