作者: Luigi D’Ambra , Biagio Simonetti , Eric J. Beh
DOI: 10.1007/978-3-7908-1709-6_21
关键词:
摘要: For the study of association in three-way, and more generally multi-way, contingency tables literature offers a large number techniques that can be considered. When there is an asymmetric dependence structure between variables Marcotorchino index [Mar84] (as apposed to Pearson chi-squared statistic) used measure strength their association. have ordinal structure, this information often not take into account. In paper we introduce partition for three ordered categorical using special class orthogonal polynomials. A graphical procedure also considered obtain visual summary asymmetrical relationship variables.