Parametric Statistical Inference

by J. K. Lindsey

Overview

Inference involves drawing conclusions about some general phenomenon from limited empirical observations in the face of random variability. In a scientific context, the general must include the completely unforeseen if all possibilities are to be considered. Many of the statistical models most used to describe such phenomena belong to one of a small number of families—the exponential, transformation, and stable families. In the past 25 years, the likelihood function has been recognized as the fundamental element of approach to drawing scientific conclusions. This book brings together for the first time these two components of statistics as the central themes of statistical inference. Chapters focus on model building, approximations, and examples. There are also appendices on the elements of measure theory, probability theory, and numerical methods. The discussions are appropriate for students of mathematical statistics.

Product Details

ISBN-13:	9780198523598
Publisher:	Oxford University Press
Publication date:	08/01/1996
Edition description:	New Edition
Pages:	512
Product dimensions:	9.21(w) x 6.14(h) x 1.13(d)

About the Author

University of Liege

Table of Contents

Part 1: Model Building1. Model building2. Exponential familyPart 2: Inference3. Likelihood4. Goodness of itPart 3: Approximations5. Asymptotics6. Factoring the likelihood functionPart 4: Decisions7. Frequentist decision-making8. Bayesian decision-makingPart 5: Examples9. Poisson regression10. Binomial regressionPart 6: AppendicesA Elements of measure theoryB Review of probability theoryC Normal distribution statisticsD Numerical methodsBibliographyIndex

Editorial Reviews

Presents statistical models belonging to the exponential, transformation, and stable families, as well as the likelihood function<-->recognized as the fundamental element of approach to drawing scientific conclusions. Includes two further chapters on approximations (asymptotics and factoring the likelihood function), and two models of human behavior in addition to statistical models<-- >the frequentist and Bayesian. The merits and problems of each are described in separate chapters. Two unifying examples bring together the various approaches to inference and decision-making. Annotation c. by Book News, Inc., Portland, Or.

Booknews