On some similarities between contingency table methods and grouped data survival analytic techniques

Abstract

Some similarities and equivalences between methods for analyzing contingency tables and techniques for analyzing grouped survival data are discussed. Attention is focused on models of homogeneity, piecewise homogeneity and models of accelerated failure rates. It is shown that under certain types of censoring, grouped data survival analytic techniques for these models are practically equivalent to methods common in contingency tables analysis with respect to models of independence, quasi-independence and uniform association models, respectively.