The confidence intervals for a ratio of two population medians are proposed and compared with the approximate confidence interval of the Price and Bonett. The proposed methods are modified from the Bonett method with the bootstrap-t method and the bootstrap percentile method. The Monte Carlo simulation technique was performed 5,000 times repeatedly, along with a round of random sampling using the bootstrap method 5,000 times. The simulation involved both data from the population with the non-normal distribution and the small size of each sample by using the R program. The results indicated that for all studied distributions, the Price and Bonett bootstrap-t confidence interval and the bootstrap percentile confidence interval performance were superior to the Price and Bonett confidence interval in 18 out of 24 case studies or 75% of case studies. The most important result of this study was that both groups showed high skewness and high kurtosis, and the Price and Bonett bootstrap-t confidence interval and the bootstrap percentile confidence interval performance were superior to the Price and Bonett confidence interval in all the studied cases.
Keywords
Bootstrap; Median; Confidence Interval; Coverage Probability; Average Length
THE JOURNAL OF KMUTNB
Published by : King Mongkut's University of Technology North Bangkok Contributions welcome at : http://www.journal.kmutnb.ac.th
By using our website, you acknowledge that you have read and understand our Cookie Policy and Privacy Policy.