An Exact Chi-Squared Decomposition Theorem for Inverse Gaussian Variates
G. Letac, V. Seshadri, G. A. Whitmore
Journal of the Royal Statistical Society: Series B (Statistical Methodology) · 1985 · Volume 47 · Issue 3
DOI: 10.1111/j.2517-6161.1985.tb01377.x
Abstract
It is proved that the scaled deviance of an inverse Gaussian sample of sizencan be expressed as a sum ofn– 1 independent chi-square variates. Applications to prediction intervals and squared recursive-residuals are illustrated.
Enable JavaScript for citation tools and additional paper interactions.