Abstract
Consider a joint variable (U 1,U 2) where U 1 and U 2 are represented as follows where Q 1,Q 2 are given positive semi definite matrices represents a chi-square random variable with νdegrees of freedom, η is a p-variate normal random vector with mean equals zero and dispersion matrix depending on a single parameter being independent. Ratio U 1/U 2 occurs in testing the equality of variances of two linear models with the same regression parameters. Here, a bivariate Gaussian approxi-mation to the distribution of is investigated through finding the distri-bution of U 1/U 2. Some other applications are also cited.

This publication has 8 references indexed in Scilit: