Premium
Simple confidence intervals for lognormal means and their differences with environmental applications
Author(s) -
Zou G. Y.,
Huo Cindy Yan,
Taleban Julia
Publication year - 2009
Publication title -
environmetrics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.68
H-Index - 58
eISSN - 1099-095X
pISSN - 1180-4009
DOI - 10.1002/env.919
Subject(s) - log normal distribution , confidence interval , statistics , coverage probability , robust confidence intervals , confidence distribution , mathematics , sample size determination , simple (philosophy) , variance (accounting) , confidence region , cdf based nonparametric confidence interval , inference , statistical inference , computer science , artificial intelligence , philosophy , accounting , epistemology , business
The lognormal distribution has frequently been applied to approximate environmental data, with inference focusing on arithmetic means. Confidence interval estimation involving lognormal means in small to moderate sample sizes has received much attention over the years without a simple procedure in sight. We therefore propose a closed‐form procedure for constructing confidence intervals for a lognormal mean and a difference between two lognormal means. The advantage of our procedure is that it only requires confidence limits for a normal mean and variance. The results of a numerical study show that our method performs as well as the generalized confidence interval (GCI) approach, which relies completely on computer simulation. Two real datasets are used to illustrate the methodology. Copyright © 2008 John Wiley & Sons, Ltd.