A New Logistic Generalization Arising from Distributions of Order Statistics: Properties and Applications

Abstract

Distributions with the variable support x∈R that exhibit strict symmetricity are versed in literature; and serve as model-fit for various forms of bell shaped outcomes; where normal and logistic distributions are renowned examples. This strictness, however, limits the application of these probability models to a particular kind of data; hence, its minimal utility. In this paper, therefore, a new generalization for the logistic distribution termed the Jones generalized logistic distribution is proposed. This new distribution is conditionally symmetric; which entails that the distribution attains symmetricity, only at equal parameter combinations. By implication, the proposed distribution serves the dual purpose of modeling both symmetric and asymmetric outcomes. Some properties of the proposed model have been derived. Finally, JGLD were fit to two different lifetime data as a demonstration to its relevance.