Wilcoxon‐type rank‐sum precedence tests: large‐sample approximation and evaluation Conferences

abstract

• AbstractWilcoxon‐type rank‐sum precedence tests are useful alternatives to the precedence life‐test proposed for testing the hypothesis that two lifetime distribution functions are equal. In this paper, we first derive the exact power function under the Lehmann alternative for these tests. We also note that the large‐sample normal approximation for the null distribution is not satisfactory in case of small or moderate sample sizes. Then, we develop an Edgeworth expansion to approximate the significance probabilities. Comparison of normal approximation and Edgeworth approximation for various sample sizes are provided. Finally, an example is presented for illustration. Copyright © 2002 John Wiley & Sons, Ltd.

authors

• Ng, HKT
• Balakrishnan, Narayanaswamy

status

• published 

publication date

• July 2002

has subject area

• 0102 Applied Mathematics (FoR)
• 0104 Statistics (FoR)
• 1502 Banking, Finance and Investment (FoR)
• Statistics & Probability (Science Metrix)

published in

• Applied Stochastic Models in Business and Industry  Journal

presented at event

• 2nd International Syymposium on business and Industrial Statistics (ISIS2)  Conference

keywords

• Edgeworth approximation
• Lehmann alternative
• Mathematics
• Mathematics, Interdisciplinary Applications
• Operations Research & Management Science
• Physical Sciences
• Science & Technology
• Statistics & Probability
• Technology
• Wilcoxon's rank-sum test
• life testing
• power
• precedence test

Digital Object Identifier (DOI)

• 10.1002/asmb.478

start page

• 271

end page

• 286

volume

• 18

issue

• 3