Tests for the Chacteristic Exponent and the Scale Parameter of Symmetric Stable Distributions
- 1 January 1990
- journal article
- research article
- Published by Taylor & Francis in Communications in Statistics - Simulation and Computation
- Vol. 19 (4) , 1465-1475
- https://doi.org/10.1080/03610919008812929
Abstract
Let S (α, β, a, c) be a stable distribution, where α∈ (0, 2), −1 ≤ β ≤ 1, a ∈ IR and c > 0 are respectively, the characteristic exponent, the skewness, the location and the scale parameters. In this paper we shall develop tests for the parameters α and c of a symmetric stable random variable S (α, 0,0, c). The test statistics for the hypthesis H0: α = α0 vs. H1: α > α0 and H0:c = c0 vs. H1 : c > c0 are functions of the empirical characteristic function , where X1, …, Xn is a random sample of size n from a stable population S (α, 0,0, c). We shall study the asymptotic properties of the power functions of the tests above and their behaviour in the neighborhood of α0 and c0.Keywords
This publication has 3 references indexed in Scilit:
- Estimation of location and scale in Cauchy distributions using the empirical characteristic functionBiometrika, 1982
- Regression-Type Estimation of the Parameters of Stable LawsJournal of the American Statistical Association, 1980
- The estimation of the parameters of the stable lawsBiometrika, 1975