Observational studies. (Q5956630)

An observational study is an empirical investigation of the effects caused by a treatment, policy, or intervention, in which it is not possible to assign subjects at random to treatments or controls, as would be done in controlled experiments. Observational studies are common in most fields that study the effects of treatments on people. This second edition is about 50 percent longer than the first one from 1995, see Zbl 0851.62081, with many new examples and methods. There are new chapters on nonadditive models for treatment effects (Chapter 5) and planning observational studies (Chapter 11). Chapter 1 defines the subject more carefully, presents several observational studies, and briefly indicates some of the issues that structure the subject. The statistical theory of randomized experiments is reviewed in Chapter 2. Chapter 3 discusses the simplest of analytical adjustments and examines the circumstances under which the adjustments succeed. Chapter 4 discusses sensitivity analyses that ask how the findings of a study might be altered by hidden biases for various magnitudes. Chapter 5 discusses models for treatment effects and illustrates their use in sensitivity analysis. Chapters 6 through 8 concern attempts to detect hidden biases using devices. Chapter 9 concerns coherence in observational studies, a concept that falls somewhere between attempts to detect hidden bias and sensitivity analyses. It has been extensively rewritten. Chapter 10 discusses methods and algorithms for matching and stratification for observed covariates. The planning of observational studies is discussed in Chapter 11. Chapter 12 discusses the relationships between the design of an observational study and its intended audience. This book is extremely useful to researchers and graduate students in the biomedical and social sciences.