We consider the problem of constructing confidence intervals for the ratio in the means of two independent populations which contain both log-normal and zero observations. We propose a maximum likelihood (ML)-based method and a two-stage bootstrap approach. We also conduct an extensive simulation study to evaluate coverage accuracy, interval width, and relative bias of the proposed methods. The simulation results indicate that when the two population skewness coefficients are the same, the ML-based interval has better coverage accuracy but is more biased than the bootstrap-based interval; when the two population skewness coefficients are different, the bootstrap-based interval has better coverage accuracy and is less biased than the ML-based interval. Finally, we analyze the charges for diagnostic tests in a study that assesses the relationship between the excess charges among older patients and the burden of their medical illness, and we find that these two are related.
ASJC Scopus subject areas
- Statistics and Probability
- Computational Mathematics
- Computational Theory and Mathematics
- Applied Mathematics