Previous |  Up |  Next


goodness-of-fit test; empirical Laplace transform; likelihood test
A sub-exponential Weibull random variable may be expressed as a quotient of a unit exponential to an independent strictly positive stable random variable. Based on this property, we propose a test for exponentiality which is consistent against Weibull and Gamma distributions with shape parameter less than unity. A comparison with other procedures is also included.
[1] Chen, Zhenmin: Statistical inference about the shape parameter of the Weibull distribution. Statist. Probab. Lett. 36 (1997), 85–90 MR 1491077 | Zbl 0916.62025
[2] Meintanis S. G.: Omnibus tests for strictly positive stable laws based on the empirical Laplace transform. Math. Meth. Statist. 14 (2005), 468–478 MR 2210542
[3] Rublík F.: Some tests on exponential populations. In: Probastat 1994, Tatra Mountains Math. Publ. 7 (1996), 229–235 MR 1408476
[4] Stehlík M.: The exact LR test of the scale in the Gamma family. In: Probastat 2002, Tatra Mountains Math. Publ. 26 (2003), 381–390 MR 2055191
[5] Stehlík M.: Exact likelihood ratio scale and homogeneity testing of some loss processes. Statist. Probab. Lett. 76 (2006), 19–26 MR 2213239 | Zbl 1085.62018
[6] Wong P. G., Wong S. P.: A curtailed test for the shape parameter of the Weibull distribution. Metrika 29 (1982), 203–209 MR 0685566 | Zbl 0492.62022
Partner of
EuDML logo