Journal of Statistics Applications & Probability
Abstract
This work aims to theoretically evaluate whether the number k of partitions of a discrete variable X affects the sensitivity Se(X) of a binary health outcome Y. The distribution of variable X was either unknown or the uniform. Thus, two discrete random variables Xk and Xk+1 with k and k+1 partitions, respectively, were considered. In addition, a random variable Y that indicates the actual health status of an individual was also considered. The case of the composite index Tkm which is formed by the sum of m variables Xk j , j = 1,2, . . . ,m either when the distribution of each variable Xk j is unknown or the uniform was also investigated. This work suggests that the sensitivity of an index is a non-decreasing function of the number of partitions, under certain conditions.
Suggested Reviewers
N/A
Digital Object Identifier (DOI)
http://dx.doi.org/10.12785/jsap/020301
Recommended Citation
Bersimis, Fragiskos; Panagiotakos, Demosthenes; and Vamvakari, Malvina
(2013)
"Sensitivity of Health Related Indices is a Non-Decreasing Function of their Partitions,"
Journal of Statistics Applications & Probability: Vol. 2:
Iss.
3, Article 1.
DOI: http://dx.doi.org/10.12785/jsap/020301
Available at:
https://digitalcommons.aaru.edu.jo/jsap/vol2/iss3/1