The course extends on Causal inference for statistics, social and biomedical sciences I. Here we will discuss causal inference using observational data. In Part III we assume that the assignment mechanism is “regular” in a well-defined sense and discuss what is called the “design” phase of an observational study. In Part IV we discuss data analysis for studies with regular assignment mechanisms. Here we consider matching and subclassification procedures, as well as model-based and weighting methods. Part V relax this regularity assumption and discuss more general assignment mechanisms. First, we assess the key unconfoundedness assumption and consider sensitivity analyses where we relax some of the key features of a regular assignment mechanism.