Pilot Project: Addressing Uncertainty in Tobacco Use Interventions

January 20, 2012Constantino Lagoa

Methodology Center Collaborative Research Funding has been awarded to Constantino Lagoa, Professor of Electrical Engineering at Penn State. Dr. Lagoa will work with Methodology Center researchers Runze Li and Linda Collins to address issues of data uncertainty in behavioral intervention design. Dr. Lagoa will draw on his background in controller design and systems modeling to inform the design of tobacco-use-reduction interventions. His objective is to help researchers design interventions that are effective for a very high percentage of participants, not just the “average” person, because people exist across a range, not clustered around an average.


In the design of behavioral interventions, researchers may know the qualitative properties of a model without being able to precisely describe the model itself. For a hypothetical (and naïve) example, we hypothesize that smoking increases with stress and that smoking, in turn, decreases stress levels. But even though we understand the relationship, we cannot describe it precisely; there is too much variability between people to describe the relationship with a single equation. In these cases, the uncertainty itself must be quantified for use in the intervention design. This project will develop robust models that account for uncertainty like this in the design of behavioral interventions.

Follow Us