Projects
Topics in post-selection inference. KU Leuven
There is an increasing awareness that care needs to be taken to properly report and incorporate all sources of uncertainty when stating research conclusions. The domain of post-selection inference concentrates on studying and developing valid inference procedures, including the construction of correct confidence intervals and providing proper p-values for hypothesis tests, when the inference takes place via a model that was not given a priori ...
Inference after model selection and averaging via confidence distributions and curves KU Leuven
Model selection and model averaging are common practices to find the best model that explains the observed data. When the working model is selected using data-driven methods and the same data are used for inference about population parameters, guarantees of classical inference techniques might not hold anymore. This dissertation discusses ways of producing valid inference for post-selection and for model averaged estimators via confidence ...
A study of correct statistical inference after model selection took place KU Leuven
Classical methods for inference in statistics assume that a model is given before looking at the data and that this model perfectly describes how the data were generated. Statistical practice proceeds in another fashion: the data are used, either visually via plots, and/or by fitting several models, performing variables selection, model selection or regularization to arrive at one or more plausible models. Those selected models are then used ...
Honest data-adaptive inference for treatment effects Ghent University
The estimation of treatment effects in observational studies is susceptible to bias due to model misspecification. Corresponding confidence intervals and p-values are moreover overly optimistic when variable selection or regularisation techniques are adopted. We will therefore develop treatment effect estimators which are less sensitive to model specification, along with honest confidence intervals and hypothesis tests.