Please use this identifier to cite or link to this item: https://hdl.handle.net/10419/300359 
Year of Publication: 
2023
Series/Report no.: 
IFS Working Papers No. 23/41
Publisher: 
Institute for Fiscal Studies (IFS), London
Abstract: 
The burgeoning use of ordinal data throughout the Empirical Sciences calls for location and variation measurement instruments suitable for such data environments. Neither Pearson's Coefficient of Variation nor the Sharpe Ratio, relative variation comparison workhorses in cardinal worlds, are applicable in ordinal paradigms without artificial data scaling, a practice recently much criticized for its inherent ambiguity. Here, employing the concept of probabilistic distance, unequivocal, scale independent, Coefficient of Variation analogues for use in Multivariate Ordered Categorical environments are introduced and exemplified in analyses of Self-Reported Health outcomes in the UK and Human Resource determinants in Canada.
Subjects: 
Ordinal Data
Coefficient of Variation
Sharpe Ratio
Probabilistic Distance
JEL: 
C18
I32
G10
Persistent Identifier of the first edition: 
Document Type: 
Working Paper

Files in This Item:
File
Size





Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.