Parametric statistical theory: with the assistance of R. Hamböker (Q1333375)

This textbook presents a survey of advanced parametric statistical theory. It requires a basic knowledge of measure theory and probability theory. Auxiliary results from measure theory and probability theory are summarized in the last section of each chapter. Chapter 1 describes the concepts of sufficiency, exhaustivity and completeness. Chapter 2 deals with the evaluation of estimators. It presents some basic definitions like unbiasedness, equivariance, concentration on intervals and on convex sets, loss function, risk and relative efficiency. Chapters 3, 4 and 5 contain exact results about confidence coefficients, significance levels, unbiasedness and optimality. Chapter 3 deals with mean unbiased estimators of minimal convex risk and chapter 4 with basic concepts of testing theory like critical region, power function, optimal tests and similar tests. Chapter 5 presents confidence procedures including optimal one-sided bounds. The rest of the book is about the asymptotic theory for the case of i.i.d. random variables. Chapter 6 investigates the consistency of estimators, especially that of \(ML\)-estimators. Chapter 7 deals with asymptotic distributions of estimator sequences and chapter 8 with asymptotic bounds for the concentration of estimators as well as with the concept of superefficiency. Chapter 9 presents miscellaneous results on asymptotic distributions. Estimates and confidence bounds for location- and scale parameters are obtained. In the last chapter 10, asymptotic power functions and bounds for the power are derived. This textbook is clear and well-written and can be highly recommended to teachers and students of mathematics. It contains a large number of exercises and examples. But, unfortunately, robust methods or at least some aspects of robustness in estimating and testing are missing, only \(M\)-estimators are introduced and their consistency is shown. However, what is the use of constructing \(M\)-estimators?