In this paper a new index to analyse the dependence between categorical variables is presented and is compared to other measures of association, mainly based on the X2 Pearson’s statistic. The new index is compared to well-known measures of cograduation, as well. To restrict our comparisons, the domain we consider includes all the square contingency tables belonging to the same Fréchet class, that is all the contingency tables in which the marginal frequencies are given as fixed in both characters. Anyway, the new index is good even if there are no constrain on the marginal distributions, but only a constrain on n, the total of the observations.
MINEO, A., RUGGIERI, M. (2005). A new index to measure association between categorical and ordinal variables. STATISTICA & APPLICAZIONI, 3 (2), 59-93.
A new index to measure association between categorical and ordinal variables
MINEO, Angelo;RUGGIERI, Mariantonietta
2005-01-01
Abstract
In this paper a new index to analyse the dependence between categorical variables is presented and is compared to other measures of association, mainly based on the X2 Pearson’s statistic. The new index is compared to well-known measures of cograduation, as well. To restrict our comparisons, the domain we consider includes all the square contingency tables belonging to the same Fréchet class, that is all the contingency tables in which the marginal frequencies are given as fixed in both characters. Anyway, the new index is good even if there are no constrain on the marginal distributions, but only a constrain on n, the total of the observations.File | Dimensione | Formato | |
---|---|---|---|
paper_Statapplic.pdf
Solo gestori archvio
Dimensione
506.01 kB
Formato
Adobe PDF
|
506.01 kB | Adobe PDF | Visualizza/Apri Richiedi una copia |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.