A New Method of Testing Log-normal Mean

In this paper, we deal with hypothesis testing for unknown parameters. A new method is proposed based on a combination of the computational approach test and maximum likelihood estimation (CAT-MLE), introduced by [7]. The CAT-MLE does not require any knowledge of the sampling distribution. It depends heavily on numerical computational and Monte-Carlo simulation study. To illustrate the methodology, we show that the CAT-MLE can be better than the two accepted tests, modified Cox method and generalized p-value method, with Monte Carlo simulations to test the mean in the lognormal distribution. We also apply our method to a set of real data.