A NEW COMPUTATIONAL APPROACH FOR TESTING EQUALITY OF INVERSE GAUSSIAN MEANS UNDER HETEROGENEITY

Authors : Esra Yigit GÖKPINAR, Esra POLAT, Fikri GOKPINAR, Süleyman GÜNAY

Pages : 581-590

View : 18 | Download : 4

Publication Date : 2013-05-01

Article Type : Research Paper

Abstract :In this article, a testing procedure based on computational approachtest is proposed for the equality of several inverse Gaussian means underheterogeneity. Not requiring the knowledge of any sampling distribution, depending heavily on numerical computations and Monte Carlosimulation, moreover, figuring out the critical region automatically arethe advantages of the computational approach test. We compare it withsome of the existing tests; the parametric bootstrap and the generalizedtest variables in terms of type I risks and powers by using Monte Carlosimulation.
Keywords : Computational approach test, Generalized test variables, Hypothesis testing, Inverse gaussian distribution, Maximum likelihood estimate, Parametric bootstrap