You are here

The Association Graph and the Multigraph for Loglinear Models

The Association Graph and the Multigraph for Loglinear Models

January 2011 | 136 pages | SAGE Publications, Inc
Though the graphical model was introduced in 1980, most of the research and application of the methodology in this field has been done by Europeans. The U.S. has lagged somewhat behind; only in recent years has the graphical model appeared in some of the American textbooks on categorical data, and even then the coverage is limited. The purpose of this work is to provide an initial source of reading for someone interested in the topic.

This publication distinguishs itself from any other book by including the multigraph representation of LLMs, a natural and very effective extension of the graphical model approach. The book's coverage would extend from the LLM, already covered by Knoke and Burke's SAGE publication, "Log-Linear Models," through the development and application of the multigraph representation in one coherent, comprehensive treatment.

About the Author
Series Editor's Introduction
Chapter 1. Introduction
Chapter 2. Structures of Association
Chapter 3. Loglinear Model Review
Chapter 4. Association Graphs for Loglinear Models
Chapter 5. Collapsibility Conditions and the Association Graph
Chapter 6. The Generator Multigraph
Chapter 7. Fundamental Conditional Independencies for Nondecomposable Loglinear Models
Chapter 8. Conclusions and Additional Examples
Data Sets
Author Index
Subject Index

Preview this book

Harry J. Khamis

Purchasing options

Please select a format:

ISBN: 9781412972383

SAGE Knowledge is the ultimate social sciences digital library for students, researchers, and faculty. Hosting more than 4,400 titles, it includes an expansive range of SAGE eBook and eReference content, including scholarly monographs, reference works, handbooks, series, professional development titles, and more.

The platform allows researchers to cross-search and seamlessly access a wide breadth of must-have SAGE book and reference content from one source.