异方差White检验应用的几个问题  

Some Application Problems of White Heteroscedasticity Test

作　　者：刘明[1,2] 

机构地区：[1]兰州商学院统计学院,甘肃兰州730020 [2]兰州商学院甘肃经济发展数量分析研究中心,甘肃兰州730020

出　　处：《统计与信息论坛》2012年第6期45-49,共5页Journal of Statistics and Information

摘　　要：White检验通常由LM检验来完成,现实中也可以使用更简单的F检验来替代LM检验;White检验过程中需要构筑辅助回归模型,它可以有不同的形式以应对实际问题的需要;讨论了四类不同的辅助回归模型,理论分析表明,它们在应用中各有所长;最后通过一个实际例子,验证了理论分析的结论,展示了White检验应用的灵活性。It is true that White test is usually done by LM test, but we also can use the F-test to replace the LM test in reality. It need to build an auxiliary regression model during the proceeding of White test, and it can take different forms in response to practical needs. It discusses four types of different auxiliary regression models in the paper, theoretical analysis showed that every of them has its own advantage. In finally we show a practical example to verify that the conclusions of the theoretical analysis, and show the flexibility of White test in application.

关 键 词：异方差 White检验 辅助回归模型 

分 类 号：O212[理学—概率论与数理统计]