Lognormal distribution - Confidence Intervals for the Mean

Abstract

Lognormal distributions are used in modelling continuous random variables which are greater than or equal to zero. A lognormal distribution can be converted into a normal distribution by logarithmic transformation of the random variable. Lognormal distributions are used in medicine to model latent periods of infectious disease, and in environmental sciences to estimate the distribution of particle size. The manuscript describes an approach to compute the confidence intervals for the mean of lognormal distribution using the loglikelihood ratio method. Simulations were used to evaluate the coverage probability for different sample sizes and the results obtained were compared with different reported methods.