| A VIEW OF THE CHI-SQUARE STATISTIC |
| |
| 引用本文： | 张尧庭,Kasra Afsarinejad. A VIEW OF THE CHI-SQUARE STATISTIC[J]. 高校应用数学学报(A辑), 1992, 0(3) |
| |
| 作者姓名： | 张尧庭 Kasra Afsarinejad |
| |
| 作者单位： | School of Management，Wuhan University，Wuhan 430072，Department of Mathematics，Linkoping University，S-581 83 Linkoping Sweden |
| |
| 摘 要： | Statistical analysis of contingency tables is essentially a discrete multivariate problem. To test independence in a two-way contingency table, one can use different principles such as general class of distance or Mahalanobis distance to derive test statistic. Most of them result in the chi-square statistic, which is a simple test statistic. The disadvantage of the chi-square test is discussed from a multivariate point of view. A new test statistic is proposed for testing independence. This statistic is more sensitive to dependence than the chi-square statistic.
|
A VIEW OF THE CHI-SQUARE STATISTIC |
| |
| Zhang Yaoting. A VIEW OF THE CHI-SQUARE STATISTIC[J]. Applied Mathematics A Journal of Chinese Universities, 1992, 0(3) |
| |
| Authors: | Zhang Yaoting |
| |
| Abstract: | Statistical analysis of contingency tables is essentially a discrete multivariate problem. To test independence in a two-way contingency table, one can use different principles such as general class of distance or Mahalanobis distance to derive test statistic. Most of them result in the chi-square statistic, which is a simple test statistic. The disadvantage of the chi-square test is discussed from a multivariate point of view. A new test statistic is proposed for testing independence. This statistic is more sensitive to dependence than the chi-square statistic. |
| |
| Keywords: | Canonical Correlation Contingency Table Entropy Independence Mahalanobis Distance |
| 本文献已被 CNKI 等数据库收录！ |
