All Times EDT
Keywords: Causal inference, mediation analysis
In randomized clinical studies, researchers want to know not only if randomized interventions work, but how they work by post-randomization mechanism or mediating variables. Especially when an intervention was not successful, researchers would like to identify possible improvements for the design and evaluation of future studies. Mediation analyses can help researchers identify, formalize, and quantify possible mechanisms (i.e., causal pathways) linking a cause to an effect. This presentation will provide a detailed introduction of mediation analyses in clinical trials with some examples.