Causal inference: from foundations to applications

4th Oct 2027
Causal inference: from foundations to applications
October 4-8 2027, La Villa Clythia – CAES du CNRS, Fréjus, France

In an era of rapidly advancing technologies and exponential data growth, many scientific and industrial challenges, such as drug discovery, public health, economic policy design, and social program evaluation, require robust methodologies to derive credible causal conclusions from both observational and experimental data. For example, a medical researcher might seek to assess the risk factors of a disease, an economist could analyze the impact of a job-training program on employment prospects, or a sociologist might study how divorce affects children’s educational outcomes.

This course focuses on discovering, identifying, and estimating causal relationships between a treatment (X) and the evolution of an outcome (Y), particularly when relying on observational data prone to selection biases and confounding factors. Building on Judea Pearl’s theory (DAGs and do-calculus) and Donald Rubin’s framework (Potential Outcomes), it introduces the fundamental principles and tools of causal inference, covering methodological, applied, computational, and mathematical aspects. This will enable participants to address real-world problems with rigor and precision.

While some experience in programming with R is a plus, it is not a prerequisite. The course will be taught in English, and no prior knowledge of causal inference is required - it will start with the basics.

The material as well as the website for the ECAS-SFdS course for 2027 are ready and the registration is open.
Details can be found at: https://sfds.asso.fr/fr/ecas2027/799-home/.