Join Us For Our Quantitative Epidemiology Methods in Practice: Seminar Series

 
Join virtually or in person!
 

MEASUREMENT ERROR IN DAGS: MISSING IN ACTION?

Presented by Melissa Wardle, MPH
PhD Student of the OHSU-PSU School of Public Health, Epidemiology

In epidemiologic research, measurement error is everywhere, but it is often missing from directed acyclic graphs (DAGs). This talk reviews the graphical structures used in DAGs and highlights the underlying epidemiological concepts they represent, taking into account measurement error. By articulating the need for incorporating measurement error into DAGs, we can make a clearer distinction between causal questions that are purely conceptual and those that represent the variables we empirically measure and use in everyday practice. Routine inclusion of measurement error in DAGs also calls into question some commonly held beliefs about the correspondence between graphical and causal structures, which will be discussed.
 
Presentation Link: https://ohsu.webex.com/ohsu/j.php?MTID=m0e58a9191165913fe2fcc74950fac443