Open Access
September, 1957 Statistical Properties of Inverse Gaussian Distributions. II
M. C. K. Tweedie
Ann. Math. Statist. 28(3): 696-705 (September, 1957). DOI: 10.1214/aoms/1177706881

Abstract

Given a fixed number n of observations on a variate x which has the Inverse Gaussian probability density function exp{ϕ2x2λ+ϕλ2x}λ2πx3,0<x<, for which E(x)=λ/ϕ=μ, it is shown how to find functions of the sample mean m whose expectations can be expressed suitably in terms of the parameter ϕ (or μ). In particular, it is shown that the conditional expectation of any unbiased estimator κ~r of the rth cumulant κr is E(κ~rm)=2m(12λn2)r1e12g1(u1)2r3e12gu2du/(r2)! where g=λn/m. This expectation may be evaluated either by series given in the paper or by using published tables of numerical values of certain functions to which it can be related. The conditional variance of the usual mean square estimator s2 of κ2 is also found. These results give an asymptotic series for the conditional variance of a generalization χs2=(n1)s2/E(s2m) of a statistic discussed by Cochran. Exact formulae for the expectation of the statistic s2/m3 and its mean square error as an estimator of λ1 are given or described. This statistic is a consistent estimator of λ1 and has asymptotically an efficiency of ϕ/(ϕ+3).

Citation

Download Citation

M. C. K. Tweedie. "Statistical Properties of Inverse Gaussian Distributions. II." Ann. Math. Statist. 28 (3) 696 - 705, September, 1957. https://doi.org/10.1214/aoms/1177706881

Information

Published: September, 1957
First available in Project Euclid: 27 April 2007

zbMATH: 0086.35202
MathSciNet: MR110132
Digital Object Identifier: 10.1214/aoms/1177706881

Rights: Copyright © 1957 Institute of Mathematical Statistics

Vol.28 • No. 3 • September, 1957
Back to Top