Ordinal association
In statistics, ordinal association is the association or relationship between two ordinal variables (for example, in a contingency table with low income, medium income, and high income in the row variable and educational level - no high school, high school, university - in the column variable).[1] There are several measures to infer the presence and direction of such relationship, such as Goodman and Kruskal's gamma, Kendall tau rank correlation coefficient, Somers' D. Some of the methods for analyzing ordinal association are common with those of rank correlation, but as in rank correlation items are in a general manner all ordered between them about a characteristic, in ordinal association items are assigned to a level in an ordinal scale, with items often having the same level and put together in a cross tabulation.
References
- ↑ Kruskal, William H. (December 1958). "Ordinal Measures of Association". Journal of the American Statistical Association. Retrieved 2012-11-04.