Uncertainty is the reason why we do evaluation research… and the reason why our research often fails to lead to change. When we use Bayesian principles to quantify uncertainty, we can help decision-makers to better understand the consequences of making decisions based on current knowledge and the potential benefits and costs of engaging in further research. In this talk, we will introduce the fundamental principles of Bayesian statistical inference and briefly talk about commonly-used statistical packages, including WinBUGS, JAGS and STAN. We will then briefly discuss applications of Bayesian inference to multi-level modeling, meta-analysis and clinical trial design. We will conclude with a preview of how Bayesian principles are being incorporated into the Multiphase Optimization Strategy (MOST) to provide a new platform for optimizing and evaluating multi-component behavioral and biobehavioral interventions.
Link below leads to an R-markdown html file with a treatment effect regression example…