Linear models. Least squares and alternatives (Q1903883)

The book reflects recent trends of research on linear statistical models as e.g. discussion of linear (deterministic and stochastic) restrictions, MDE-comparisons of estimators, prediction, sensitivity analysis, incomplete data sets, robust regression and binary response variables. In an appendix one finds useful results in matrix algebra. (Unfortunately, in the text some of the references to the appendix are incorrect). The theory is illustrated by well chosen examples (there are no exercises), although I felt uneasy with the introductory example, the whole chapter 2. This chapter contains a flood of econometric terms which will never be used later. Also the introduction of the data as a stochastic process is not typical for the rest of the book. The book is partially in definition - theorem - proof style (for more complicated results there are references to the literature), partially in informal style with only few motivations or discussions of the purely statistical background. It seems hard to me to specify which previous knowledge is assumed. In spite of the high expertise both authors have in this area I think that other competing books are better organized and more suitable as text books.