The Linear Model and Hypothesis: A General Unifying Theory

ISBN-13: 9783319349176 / Angielski / Miękka / 2016 / 205 str.

George Seber

Recenzje

"The book deals with the classical topic of multivariate linear models. ... the monograph is a consistent, logical and comprehensive treatment of the theory of linear models aimed at scientists who already have a good knowledge of the subject and are well trained in application of matrix algebra." (Jurgita Markeviciute, zbMATH 1371.62002, 2017)

"This monograph is a welcome update of the author's 1966 book. It contains a wealth of material and will be of interest to graduate students, teachers, and researchers familiar with the 1966 book." (William I. Notz, Mathematical Reviews, June, 2016)

Spis treści

1.Preliminaries.- 2. The Linear Hypothesis.- 3.Estimation.- 4.Hypothesis Testing.- 5.Inference Properties.- 6.Testing Several Hypotheses.- 7.Enlarging the Model.- 8.Nonlinear Regression Models.- 9.Multivariate Models.- 10.Large Sample Theory: Constraint-Equation Hypotheses.- 11.Large Sample Theory: Freedom-Equation Hypotheses.- 12.Multinomial Distribution.- Appendix.- Index.

O autorze

George Seber is an Emeritus Professor of Statistics at Auckland University, New Zealand. He is an elected Fellow of the Royal Society of New Zealand, recipient of their Hector medal in Information Science, and recipient of an international Distinguished Statistical Ecologist Award. He has authored or coauthored 16 books and 90 research articles on a wide variety of topics including linear and nonlinear models, multivariate analysis, matrix theory for statisticians, large sample theory, adaptive sampling, genetics, epidemiology, and statistical ecology.

Blurb

This book provides a concise and integrated overview of hypothesis testing in four important subject areas, namely linear and nonlinear models, multivariate analysis, and large sample theory. The approach used is a geometrical one based on the concept of projections and their associated idempotent matrices, thus largely avoiding the need to involve matrix ranks. It is shown that all the hypotheses encountered are either linear or asymptotically linear, and that all the underlying models used are either exactly or asymptotically linear normal models. This equivalence can be used, for example, to extend the concept of orthogonality in the analysis of variance to other models, and to show that the asymptotic equivalence of the likelihood ratio, Wald, and Score (Lagrange Multiplier) hypothesis tests generally applies.