Applying Mediation Analysis To Understand How Interventions Work

Medicine: Mind the Gap Seminar Dr. MacKinnon will describe mediation analysis methods with attention to solutions for some of the limitations of these methods. Mediating variables have a long and important history in theoretical and applied research, because they describe how and why two variables are related. One common example is the study of the mediating processes that explain how a prevention or treatment program achieves its effects on an outcome variable. If the active ingredients are identified, then interventions can be made more powerful and more efficient. Other mediating variable examples include how a risk factor leads to disease and how early life experiences affect later development. Over the last decade, there has been considerable development of new methods and wider substantive application of mediation analysis. The presentation will describe the questions mediation analysis can answer and how mediating variables differ from moderators, confounders, and covariates. The analysis of the single- and multiple-mediator models are used to demonstrate several controversial issues in inference, sigficance testing, and confidence interval estimation. New models for longitudinal mediation and approaches to investigating assumptions of the mediation model are described. Dr. MacKinnon will also discuss future directions in mediation theory and statistical analysis.For more information go to https://prevention.nih.gov/programs-events/medicine-mind-the-gapAir date: 4/8/2...
Source: Videocast - All Events - Category: Journals (General) Tags: Upcoming Events Source Type: video