TL;DR: In this article, an extended growth curve model is considered, which is useful when linear restrictions exist on the mean in the ordinary growth curve, and the maximum likelihood estimators consist of complicated stochastic expressions.
TL;DR: In this paper, the moments and the p.d.f. of some test criteria, namely Wilks Wp and Wilks-Laweley Up, were derived in connection with tests: (I) equality of two covariance matrices of two complex normal distributions and (II) p-dimensional mean vectors of l p-variate normal populations.
Abstract: Let S 1 and S 2 be independently distributed having respectively the complex noncentral Wishart and the central Wishart . Based on the density function of S 1 S 2 -1 under violation, the moments and the p.d.f. of some test criteria, namely Wilks Wp and Wilks-Laweley Up are derived in connection with tests: (I) Equality of two covariance matrices of two complex normal distributions and (II) Equality of p-dimensional mean vectors of l p-variate normal populations. The results are useful for drawing inferences about the robustness of test criteria concerning test of (I) when the assumption of normality is violated and of (II) when that common covariance matrix is disturbed. The moments and density function 0f tests associated with the classical testing of (I)and(II) may be obtained as special case of the results furnished here, particularly when