In a two-way contingency table, the analyst is mast interested in the hypotheses of either homogeneity or independence. For testing this as a null hypothesis, Pearson's x^(2) statistic is most commonly used in practice. Once the null hypothesis is rej...
In a two-way contingency table, the analyst is mast interested in the hypotheses of either homogeneity or independence. For testing this as a null hypothesis, Pearson's x^(2) statistic is most commonly used in practice. Once the null hypothesis is rejected, he will further search for cells which caused the rejection of the null hypothesis. For this purpose, so called cell x^(2) components are used. Kim and Lee (1996) derived the influence of an observation to the x^(2) statistic as a function. Kim (1998) derived the influence function of a cell to the x^(2) statistic. In this paper, the result of these papers is extended to derive the row influence function to the x^(2) statistic. A numerical example is given to demonstrate the role of the new function.