The need for building and generating statistically dependent random variables arises in various fields of study where simulation has proven to be a useful tool. In this work, we present an approach for constructing ordinal variables with arbitrarily assigned marginal distributions and value of association or correlation, expressed in terms of either Goodman and Kruskal's gamma or Pearson's linear correlation. The approach first constructs a class of bivariate copula-based distributions matching the assigned margins, and then, within this class, identifies the distribution matching the assigned association or correlation, by calibrating the copula parameter. A numerical example and a possible application are illustrated.
Inducing a Target Association between Ordinal Variables by Using a Parametric Copula Family / A. Barbiero. - In: AUSTRIAN JOURNAL OF STATISTICS. - ISSN 1026-597X. - 49:4(2020 Apr 14), pp. 9-18. [10.17713/ajs.v49i4.1116]
Inducing a Target Association between Ordinal Variables by Using a Parametric Copula Family
A. Barbiero
Primo
2020
Abstract
The need for building and generating statistically dependent random variables arises in various fields of study where simulation has proven to be a useful tool. In this work, we present an approach for constructing ordinal variables with arbitrarily assigned marginal distributions and value of association or correlation, expressed in terms of either Goodman and Kruskal's gamma or Pearson's linear correlation. The approach first constructs a class of bivariate copula-based distributions matching the assigned margins, and then, within this class, identifies the distribution matching the assigned association or correlation, by calibrating the copula parameter. A numerical example and a possible application are illustrated.File | Dimensione | Formato | |
---|---|---|---|
1116-Article Text-4317-1-10-20200413.pdf
accesso aperto
Descrizione: Publisher's version
Tipologia:
Publisher's version/PDF
Dimensione
264.52 kB
Formato
Adobe PDF
|
264.52 kB | Adobe PDF | Visualizza/Apri |
Barbiero_AJS - revised.pdf
accesso aperto
Descrizione: version accepted by Publisher
Tipologia:
Publisher's version/PDF
Dimensione
299.89 kB
Formato
Adobe PDF
|
299.89 kB | Adobe PDF | Visualizza/Apri |
Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.