A Comprehensive Review of the Two-Sample Independent or Paired Binary Data, with or without Stratum Effects

Abstract

Various statistical hypotheses testing for discrete or categorical or binary data have been extensively discussed in the literature. A comprehensive review is given for the two-sample binary or categorical data testing methods on data with or without Stratum Effects. The review includes traditional methods such as Fisher’s Exact, Pearson’s Chi-Square, McNemar, Bowker, Stuart-Maxwell, Breslow-Day and, Cochran-Mantel-Haenszel, as well as newly developed ones. We also provide the roadmap, in a figure or diagram format to which methods are available in the literature. In addition, the implementation of these methods in popular statistical software packages such as SAS and/or R is also presented. This will be helpful for researchers to determine which (categorical-data) testing method is available to use in various fields of study such as clinical trials, epidemiology, etc., both for the design phase of a study in prospective study, cross-sectional or retrospective study analysis.