Theory Seminar: Causal Graphs: An Introduction

Johannes Knörzer & Patrick Emonts (MPQ)
Modeling causality through directed acyclic graphs (DAGs) offers an appropriate language to think about and describe causal relationships between events.

Johannes Knörzer & Patrick Emonts (MPQ)
Herbert-Walther Lecture Hall G0.25
Wed, 24. February 2020, 11:30 am

Abstract:

Modeling causality through directed acyclic graphs (DAGs) offers an appropriate language to think about and describe causal relationships between events. In this talk, we will talk about graphical models and how they are used in causal reasoning. We will explain useful concepts such as d-separation, the back-door criterion and do calculus. To demonstrate the usefulness of this framework, we are going to provide specific examples, with data taken from real studies.