A0191
Title: Causal inference with functional data
Authors: Dominik Liebl - University Bonn (Germany) [presenting]
Tim Mensinger - University of Bonn (Germany)
Abstract: Building upon the potential outcome framework for functional data, a method is proposed for inferring treatment effect functions in function-on-scalar regression models using simultaneous confidence bands based on adaptive critical value functions. The proposed method allows for inference under fairness constraints, enabling both local and global interpretability of the results. We present convergence and asymptotic normality results for the functional version of the doubly-robust augmented inverse propensity score estimator.