For two growth curve models g1 = G (X1BZ1, V1, In1 ) and g2 = G(X2BZ2, V2,In2 ), where V1 and V2 are known symmetric nonnegative definite matrices, we mfore a com-parison between them in estimable subspace D and obtain several necessary and sufficientconditions of g1 g2(D).
This paper studies the Potthoff-Roy transformation in growth curve models and obtains the best Potthoff-Roy transformation under the criterion of minimum matrix norm. The paper proposes further an approach to improve the Potthoff-Roy transformation by using the covariates. The selection of covariates and its influence on the estimation of parameters in also studied.
In this paper, the growth data of body length and weight of cattle in the southern areas and mountainous districts of Fujian were fitted in with the eight growth curve models which were built and discussed in another paper.
The problems on testing position and scale parameters together of a growth cure model in complte case are considered in this paper. let the null hypothes be H Σ=I and ξ=0, asympstotic non - null distributions of the modified likelihood ratio statistic under two alternatives which dose to the null hypothesis are given respectively.
Die vorliegende Arbeit unternimmt den Versuch, die "Speed"-Hypothese unter Verwendung von "Latent Growth Curve Models" auf ihre individuelle Gültigkeit für die Entwicklung der fluiden Intelligenz hin zu überprüfen.
Linear and quadratic growth curve models with intraclass covariance structure and related optimal designs
It is shown that for exponential growth curve models with one parameter, maximin efficient designs can not be one point designs.
A similar result is obtained for growth curve models with two parameters.
Asymptotically optimal tests for some growth curve models under non-normal error structure
This paper gives the necessary and sufficient(n.s.) condition for the existence of Gauss-Markov Estimator(GME) of Eyy′ in the following Growth Curues Model Ey=X_1BX′_2, covy=Σ(?)V, and gives the n.s.condition for existence of GME of Ey.Where Y_(n×p)=(y_1, …,y_p)·y=(y′_1,…,y′_p)′is a quasinormally distributed random vector.
For the growth curve model, a method of calculating non-negative estimator of the parametric matrix is given in this paper,
The problems on testing scale parameter is Growth. Curve Model are considered in this paper. The asymptotic non-null dis-tributions of tie modified likelihood ratio statistic under three alterna-tives are given respectively.