How to draw DAGs (from expert knowledge)

The minimal requirements for getting domain experts drawing a causal model along with guidelines for facilitating. We will distinguish what an expert participant needs to know with what the modelling facilitator needs to know, and provide key lessons learned from running these elicitation sessions with domain experts. 

How to draw DAGs (from data)

Working through a causal discovery example, including how to support this with domain knowledge. This will follow the code in this github repository, and offer example code in R and Python. 

Why: the implications of a causal DAG

Using the underlying theory, supported by the DAGitty tool, to explore what the implications and affordances of the DAG are.

Experimental design reflected in DAGs

How DAGs can be leveraged to make causal claims from both experimental and observational data.

Where to next? 

Planning future research collaboration with workshop participants. This may cover working with similar research problems to build common models, the potential for a repository of models of common interest, or opportunities for special issue publications.

We will collect ideas, contact lists, and potential collaboration projects on the Future collaboration page, which will open closer to the conference beginning.