Chapter 10. Models for Matched Pairs
Published Online: 26 MAR 2003
DOI: 10.1002/0471249688.ch10
Copyright © 2002 John Wiley & Sons, Inc.
Book Title
Categorical Data Analysis, Second Edition
Additional Information
How to Cite
Agresti, A. (2003) Models for Matched Pairs, in Categorical Data Analysis, Second Edition, John Wiley & Sons, Inc., Hoboken, NJ, USA. doi: 10.1002/0471249688.ch10
Publication History
- Published Online: 26 MAR 2003
- Published Print: 3 JUL 2002
Book Series:
ISBN Information
Print ISBN: 9780471360933
Online ISBN: 9780471249689
- Summary
- Chapter
Keywords:
- matched pairs;
- McNemar test;
- conditional logistic regression;
- case-control studies;
- square tables;
- quasi symmetry;
- quasi independence;
- agreement;
- kappa;
- Bradley–Terry model
Summary
Chapter 10 is the first of four chapters dealing with forms of repeated measurement and other types of clustered data. It focuses on models for matched pairs. It introduces the McNemar test, and shows connections with conditional logistic regression for binary matched pairs. It then introduces marginal models for square contingency tables, and loglinear models for such tables such as quasi symmetry and quasi independence. It presents applications for square tables, such as measuring agreement between observers using kappa and using models, and the Bradley–Terry model for paired preferences.