DIAGNOSTICS FOR ASSOCIATION MODELS FOR THE ANALYSIS OF CROSS-CLASSIFICATIONS HAVING ORDERED CATEGORIES

DOI Web Site

Bibliographic Information

Other Title
  • Diagnostic for Association Models for the Analysis of Cross-Classifications Having Ordered Categories
  • Diagnostic for Association Models for t

Search this article

Abstract

This article describes some diagnostic tools for association models in the analysis of cross-classifications having ordered categories. A general technique for assessing the outlying and influential cells in two-way contingency tables is discussed. These cells can be sequentially identified using diagnostic measures (e. g., the residual from the expected frequency with the cell deleted), and tested in terms of the reduction in the value of the likelihood ratio goodness-of-fit statistic. The influence due to modify the original table, for example, with columns interchanged and/or rows combined can be investigated by diagnostic measures. The influence by treatment of sampling zero entries in contingency tables is also examined. The utility of the methods developed in this article is demonstrated with the analysis of four examples, and new insights into these data are obtained.

Journal

Keywords

Details 詳細情報について

Report a problem

Back to top