Directed Acyclic Graphs (DAGs): Introduction and applications.

The Environmental Epidemiology for Risk Assessment Committee will be hosting a webinar by Dr. Hsin-Yi Weng titled "Directed Acyclic Graphs (DAGs): Introduction and applications".

Click here to view the webinar slides

Click here to watch the webinar recording

Abstract
Causal assumptions are required in confounding evaluation. Directed acyclic graphs (DAGs) can be used to communicate underlying causal assumptions and inform variable selection for statistical adjustment. This webinar will start with a brief introduction of DAG terminology followed by examples to illustrate the utilization of DAGs in confounder identification as well as the application of DAGs in environmental epidemiology. Advantages of DAGs over traditional approaches for confounder adjustment will be discussed. Finally, a hands-on demonstration of an online program (DAGitty) for construction and analysis of DAGs will be presented.

Speaker Bio
Hsin-Yi Weng, BVM, MPH, PhD, is an Associate Professor of Clinical/Analytical Epidemiology at the Department of Comparative Pathobiology, College of Veterinary Medicine, Purdue University. Her research interests are in applying epidemiologic methods to studies promoting healthy, positive, and safe human-animal interactions in vulnerable populations (such as pregnant women, new mothers, and young children). As in studies of assessing harmful effects of environmental exposures, very often observational studies are the only ethically feasible options to quantify health effects of family pets. Epidemiological reasoning becomes essential in making valid inference from observational studies.