Bootstrap tests for independence in two-way ordinal contingency tables

摘要

For the analysis of an r×c contingency table having ordered row categories and ordered column categories, a bootstrap method is applied for the model-based likelihood ratio test for independence. A model-based likelihood ratio chi-square statistic and the statistic of the maximum eigenvalue of a Wishart matrix are also discussed. A simulation study is performed to compare the proposed method with existing ones. A real data example is included.