一类分块混合效应模型的参数估计

研究一类分块混合效模型中部分固定效应向量的估计问题，为得到其可行的最佳线性无偏估计，提出了一种减约模型，并给出了由此减约模型得出的可行估计等于原模型下相应的最佳线性无偏估计的充分必要条件，并以实例说明文中结果。

Parameter Estimation in a Partitioned Mixed-Effects Model

For a class of partitioned mixed-effects linear model, a reduced model is introduced in order to obtain the best feasibl e linear unbiased estimator of a sub-vector of fixed effects. Some sufficient a nd necessary conditions for the estimators under the reduced model to be equal t o the best linear unbiased estimte under the original model are obtained. To ill ustrate theoretical results, a practical example is also given.