摘要
In this paper, we consider whether the random effect exists in linear mixed models （LMMs） when only moment conditions are assumed. Based on the estimators of parameters and their asymptotic properties, a Wald-type test is constructed. It is consistent against global alternatives and is sensitive to the local alternatives converging to the null hypothesis at parametric rates, a fastest possibly rate for goodness-of-fit testing. Moreover, a simulation study shows the performance of the test is good. The procedure also applies to a real data.
In this paper, we consider whether the random effect exists in linear mixed models （LMMs） when only moment conditions are assumed. Based on the estimators of parameters and their asymptotic properties, a Wald-type test is constructed. It is consistent against global alternatives and is sensitive to the local alternatives converging to the null hypothesis at parametric rates, a fastest possibly rate for goodness-of-fit testing. Moreover, a simulation study shows the performance of the test is good. The procedure also applies to a real data.
基金
Supported by a grant (HKBU2030/07P) from the Research Grants Council of Hong Kong
the National Natural Science Foundation of China (Grant No. 10871001)
the Humanities and Social Sciences Project of Chinese Ministry of Education (Grant No. 08JC910002)
Zhejiang Provincial Natural Science Foundation of China (Grant No. Y6090172)
Youth Talent Foundation of Zhejiang Gongshang University, China
