A form of bivariate binomial conditionals distributions

Abstract.

In this article, we discuss a bivariate distribution whose conditionals are univariate binomial distributions, but the marginals are not binomial that exhibits both positive and negative correlation. Some useful structural properties of this distribution namely marginals, moments, generating functions, stochastic ordering are investigated. Simple proofs of both positive and negative correlation, marginal over-dispersion, the distribution of the maximum and minimum order statistics for a random sample of size 2 are also derived. The distribution is shown to be a member of the multi-parameter exponential family, and some natural and useful consequences are also outlined. The proposed distribution reduces to another bivariate discrete distribution, namely the bivariate Poisson conditionals recently studied by Ghosh, Marques, and Chakraborty (2021) under certain parametric conditions. Finally, the distribution is fitted to two bivariate count data sets having positive and negative correlation separately to illustrate its’ suitability.

Keywords:

• Bivariate binomial distribution
• conditional specification
• negative and positive correlation
• conditional failure rate
• limiting distribution

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The work of Filipe J. Marques is funded by national funds through the FCT – Fundaç∼ao o para a Ciência e a Tecnologia, I.P., under the scope of the projects UIDB/00297/2020 and UIDP/00297/2020 (Center for Mathematics and Applications).