Interval estimation of Universal Variation Ratio

Abstract

Qualitative variables are of great importance in the social sciences and descriptive measures of variability and shape have been developed for these variables, but they are little used. One of the main reasons is that they are not available in statistical software. This paper focuses on a recently introduced measure of variability, the Universal Variation Ratio, which enhances upon variation ratios previously formulated. This article aims to propose the calculation of its standard error and confidence interval and to provide a script for its computation using the R program. It applies to two scenarios: with one mode and with two or more modes. It is recommended to integrate this variability measure into data analysis procedures.