• A STUDY ON TWO PARAMETER DISCRETE QUASI LINDLEY DISTRIBUTION AND ITS DERIVED DISTRIBUTIONS

Abstract

Modeling count data is one of the most important issues in statistical research. In this paper, a new probability mass function is introduced by discretizing the continuous failure model of the Quasi- Lindley distribution. The discrete Quasi Lindley (DQL) distribution has been derived and further certain properties of the distribution have been discussed. Properties such as the recurrence relations for probabilities, factorial moments and index of dispersion of this distribution are also investigated. Estimation of parameters of DQL distributions have been discussed. The size- biased, Zero- truncated and Zero- modified forms of DQL distribution have also been investigated. To test its goodness of fit, DQL distributions have been fitted to some of well known data sets where discrete Poisson- Lindley distribution and discrete gamma distributions have earlier been fitted by others. The results show that, the two parameter DQL distribution can provide a better fit than the other derived distributions. It is noted that DL is a particular case of DQL distribution.