数学物理学报

• 论文 • 上一篇    下一篇

半参数随机效应模型的异方差检验

朱仲义;冉昊   

  1. 复旦大学统计系 上海 200433
  • 收稿日期:2004-09-08 修回日期:2005-10-09
  • 通讯作者: 朱仲义
  • 基金资助:
    国家自然科学基金(10371042)和上海市科委重点基金(02DJ14063)资助

Testing for Heteroscedasticity in Semiparametric
Random Effect Model

Zhu Zhongyi ;Ran Hao   

  1. Department of Statistics, Fudan University, Shanghai 200433
  • Received:2004-09-08 Revised:2005-10-09
  • Contact: Zhu Zhongyi

摘要: 在许多实际问题中,检验观察数据是否出现异方差性是一个相当感兴趣的问题.该文研究了半参数随机效应模型的异方差检验问题.基于Lin(1997)的方法,得到了检验方差成分都为零的Score检验统计量.通过随机模拟和实际数值例子,论证了方法的有效性.利用现有的统计软件,容易实现该文所提出的检验方法.

关键词: 异方差, 随机效应, Score检验, 半参数回归, 光滑样条

Abstract: It is of considerable interest in testing for heteroscedasticity in many practical studies. In this paper, the authors discuss this type of problem in framework of semiparametric mixed models. A global score test is proposed for the null hypothesis that all the variance components are zero. The test is based on the approaches of Lin (1997), the authors examine its performance in a simulation study, illustrate the test methods by using the progesterone data. The test can be easily implemented by using existing statistical softwares.

Key words: Heteroscedasticity, Random effects, Score test, Semiparametric regression models, Smoothing spline

中图分类号: 

  • 62G08