No label defined (Q3217449)
From MaRDI portal
!
This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:
scientific article; zbMATH DE number 3883421
| Language | Label | Description | Also known as |
|---|---|---|---|
| default for all languages | No label defined |
||
| English | No label defined |
scientific article; zbMATH DE number 3883421 |
Statements
1983
0 references
worked examples
0 references
discussion of computer packages
0 references
likelihood ratio approach
0 references
tests of significance
0 references
Roy's union-intersection principle
0 references
Bonferoni's inequalities
0 references
confidence intervals
0 references
SAS matrix procedures
0 references
t- tests
0 references
analysis of variance
0 references
multiple regression
0 references
tests for equality of covariance and independence
0 references
discriminant analysis
0 references
dimension-reducing procedures
0 references
canonical correlation
0 references
incomplete data
0 references
tests for shift in the mean
0 references
0.8563822507858276
0 references
Our emphasis is on methods in current use in multivariate statistics. For each new topic we present not only a description of the problem and its solution, but also several worked examples, chosen from many different fields. Each chapter ends with a discussion of computer packages available and additional worked examples. The likelihood ratio approach has been adopted for tests of significance of a given hypothesis, and Roy's union-intersection principle and Bonferoni's inequalities for confidence intervals are introduced. For each test statistic, formulas for the observed significance levels are given. The percentage points needed to calculate confidence intervals appear as Appendixes. Chapters 1 and 2 provide a review of the necessary matrix theory and statistical theory; Chapter 1 also contains a discussion of the SAS matrix procedures for calculating eigenvalues, eigenvectors, and the generalized inverse of a matrix. Chapters 3-7 are multivariate generalizations of univariate procedures for t-tests, analysis of variance, and multiple regression. Chapter 12 gives tests of the assumptions required for these multivariate procedures to be valid, including tests for equality of covariance and independence. It is Chapters 8-11 that contain strictly multivariate procedures, beginning with discriminant analysis. Methods for finding functions that will discriminate among populations or groups are discussed, and procedures to reduce the number of characteristics necessary for discrimination are given in a section on stepwise discriminant analysis. Chapters 9-11 then introduce dimension-reducing procedures, including canonical correlation, which investigates the correlation between linear combinations of variables, and principal component analysis, which reduces the set of measured characteristics to fewer components. Advanced topics are given toward the end of a chapter. For example, Chapter 3 covers the problem of incomplete data and tests for shift in the mean. The core material may be used in a one-semester course in applied multivariate statistics at the senior or first-year graduate level. Two semesters are required to cover the text in its entirety.
0 references