The Analysis of Cross-Classification Data: A Prediction Approach.

This paper introduces a categorical data analysis technique called Hildebrand's del that allows for the testing of customized prediction rules, and provides a strength of a strength in categorical prediction rules.

Abstract

This paper introduces a categorical data analysis technique called Hildebrand's del. The advantages of del are; it allows for the testing of customized prediction rules, it provides a strength of association measure, and, it is independent of sample size. The del procedure is illustrated via longitudinal data analysis.