Comparative study on methods of generating random variables from the power Lindley distribution

Abstract

In this paper we considered some methods of generating random variables from the power Lindley distribution using the Composition, Lambert W function and Numerical methods. In order to compare these methods with respect to their ability to generate random samples from the distribution, a Monte Carlo simulation study was conducted to show the performance and asymptotic behavior of the maximum likelihood estimates (MLEs) of the power Lindley distribution. Three quantities such as the average bias, average mean square error and coverage probability of 95% confidence interval for the parameter estimates were computed for each method and the results show that the Lambert W function and the Numerical methods which gave same result are more efficient in generating random samples from the Power Lindley Distribution than the one based on composition method.