Proportionaliy of covariance matrices (Q1093288)

S\({}_ 0,S_ 1,...,S_ k\) are mutually independent p by p matrices, \(S_ i\) having a Wishart distribution with \(n_ i\) degrees of freedom and expectation \(\Sigma_ i\). The likelihood ratio test of the hypothesis \(\Sigma_ i=\lambda_ i\Sigma_ 0\) for \(i=1,...,k\) is developed. Assuming the hypothesis is true, it is shown that the maximum likelihood estimators of \(\Sigma_ 0,\lambda_ 1,...,\lambda_ k\) are given by the unique solution to the likelihood equation, and an iterative procedure is given for the computation of the maximum likelihood estimators. The distribution of the maximum likelihood estimators is approximated.