Asymptotic properties of the growth curve model with covariance components (Q1265610): Difference between revisions

In a multivariate regression (growth curve model) model the covariance matrix of the observation matrix is considered in the form of a tensor product. The first term in this product is a known matrix, the other term is a linear combination of known matrices and unknown parameters (covariance components). The explicit form of the locally best estimators of the covariance components under normality and asymptotic confidence ellipsoids for certain linear functions of the first order parameters (parameters of the mean value of the observation matrix) estimating simultaneously the first and the second order parameters are derived in the case of replicated observations.