Abstract
The Inverse Gaussian (IG) distribution is widely used in practice and therefore an important issue is to develop a powerful goodness-of-fit test (GOF) for this distribution. In this article, we propose and examine a new GOF test for the IG distribution based on a new estimate of Kullback-Leibler (KL) information. The properties of the test statistic are presented. In order to compute the proposed test statistic, parameters of the IG distribution are estimated by maximum likelihood estimators, which are simple explicit estimators. Critical values and the actual sizes of the proposed test are obtained. Through a simulation study, power values of the proposed test are compared with some prominent existing tests. Finally, two illustrative examples are presented and analyzed.