Title: Inference for factorial designs in survival models
Authors: Markus Pauly - Technical University of Dortmund (Germany)
Arnold Janssen - Heinrich-Heine University Duesseldorf (Germany)
Marc Ditzhaus - Technical University of Dortmund (Germany) [presenting]
Abstract: Inference procedures are proposed for general nonparametric factorial survival designs with possibly right-censored or left-truncated time-to-event data. Similar to additive models, null hypotheses are formulated in terms of cumulative hazards. Thereby, deviations are measured in terms of quadratic forms in Nelson-Aalen-type integrals. In contrast to existing approaches, this allows us to work without restrictive model assumptions on the survival, censoring or truncation times. For a distribution-free application of the method, permutation and wild bootstrapping are suggested. The resulting procedures' asymptotic validity, as well as their consistency, are proven, and their small sample performances are analyzed in extensive simulations. Finally, their applicability is demonstrated by an illustrative data analysis.