While the adage “correlation does not imply causation” is true, work in the area of causal inference has pushed understanding of what kinds of causal claims can be made, especially in the context of observational data.
This workshop, the first of two, will provide an overview of causal inference. Topics will include causal diagrams and causal paths, using the tool of directed acyclic graphs (DAGs) to inform choices such as when variables should or shouldn’t be controlled for. Examples will be provided using the R software package, but no knowledge of R will be needed.