Rank correlation is a fundamental tool for expressing dependence in cases where data arearranged in order. Measures of rank correlation have been developed in several contexts formore than a century, from simple ones to relatively complicated denitions involving oneor more systems of weights. However, only a few of these can actually be considered to bereasonable measures of concordance/discordance. The main drawback with the vast majorityof coecients is their esistance-to-change", which appears to be of limited value for thepurposes of rank comparisons that are intrinsically robust. In this article, a new correlationmeasure is dened that is based on the ratios of ranks and anti-ranks. In comparing it withexisting rank correlations, we found it to be extremely sensitive to permutation patterns. Wehave illustrated the potential improvement that our index can provide in economic contextsby comparing published results with those obtained through the use of this new index. Thesuccess that we have had suggests that our index may have important applications whereverthe discriminatory power of the rank correlation coecient should be particularly strong.

A coefficient of correlation based on ratios of ranks and anti-ranks

TARSITANO, Agostino
;
LOMBARDO, Rosetta
2013

Abstract

Rank correlation is a fundamental tool for expressing dependence in cases where data arearranged in order. Measures of rank correlation have been developed in several contexts formore than a century, from simple ones to relatively complicated denitions involving oneor more systems of weights. However, only a few of these can actually be considered to bereasonable measures of concordance/discordance. The main drawback with the vast majorityof coecients is their esistance-to-change", which appears to be of limited value for thepurposes of rank comparisons that are intrinsically robust. In this article, a new correlationmeasure is dened that is based on the ratios of ranks and anti-ranks. In comparing it withexisting rank correlations, we found it to be extremely sensitive to permutation patterns. Wehave illustrated the potential improvement that our index can provide in economic contextsby comparing published results with those obtained through the use of this new index. Thesuccess that we have had suggests that our index may have important applications whereverthe discriminatory power of the rank correlation coecient should be particularly strong.
Ordinal data; Nonparametric agreement; Economic applications of ranks
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/20.500.11770/146170
 Attenzione

Attenzione! I dati visualizzati non sono stati sottoposti a validazione da parte dell'ateneo

Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus ND
  • ???jsp.display-item.citation.isi??? 2
social impact