A Novel Survival Class NBULC of Life Distributions: Statistical Testing and Applications to Scientific Data

Abstract

This paper discusses a novel approach by examining the failure patterns in the recorded survival data. Following the use of the suggested strategy, survival data are recorded. It is examined to be based on a constant failure rate (null hypothesis) or the new class called new better than used in the Laplace transform of convex order (NBULC) class of life distribution (alternative hypothesis); then the data in use provided a better or higher total present value than an older component in convex (positive or negative effects). The proposed class of life distribution NBULC included many classes, like NBU (new better than used) and NBUL (new better than in Laplace transform) classes of life distributions. The suggested test statistics for our class of life distribution is based on the goodness-of-fit method for non-censored and censored samples. The distribution of this test statistic is investigated via a simulation study. Scientific data is considered an application that utilizes real test data.