Application deadline is September 25. Start date is flexible, preferably before February 1st.

You will join the “Safe Causal Inference” consortium. Causal inference studies what assumptions are needed, and which methods you can use when those assumptions are met, to reliably conclude that “A causes B” and not just “A is correlated with B”.

You will work on the subproject “Triangulation of performance assessment methods for conditional causal quantities”. This PhD project investigates how to reliably assess the performance of estimators of conditional causal effects (such as CATE estimators and predictions under interventions). The core problem is that evaluating an estimator requires counterfactual outcomes in the evaluation dataset, but deriving those outcomes itself relies on causal assumptions, meaning the evaluation can be just as assumption-dependent as the estimator it’s meant to assess.

More information: https://workingat.vu.nl/vacancies/phd-postion-in-area-of-causal-inference-amsterdam-1335952