About the group
Causal inference is the formal process of drawing conclusions from data about the effects of treatments or other types of interventions. Our aim is to estimate treatment effects that are simple to interpret, and to avoid bias in estimated effects due to spurious associations. This involves proper specification of target parameters, establishing conditions that allow identification of such effects, and in the end, estimation.
The area of causal inference has over the last decades grown to be a very active area within statistics. Various new methods have been introduced, and the aim of this group is to contribute to the development in this field. We focus on methods for causal inference in survival and event history analysis, which are typically complicated by censoring and the fact that treatments, outcomes and other key variables, are processes that change over time. Some examples are methods for estimating the effects of time-varying treatments such as marginal structural models, continuous time models, mediation analysis, competing risk and multi-state outcomes, etc.