摘要
利用回归模型进行统计控制,在实际工作中一直得到广泛应用。通常采用的方法是利用回归模型y^=a+bx进行逆估计,即根据回归模型的变换x=(y^-a)/b,由应变量Y的取值范围反推自变量X的取值范围。本文指出这种方法是很不合理的;进一步提出二种相对合理取代的统计方法。
Statistical control in regression model has been applied widely in practice, It is an usual method to estimate reversely in regression model y^ = a ＋ bx, That is to say, according to the inverse form of regression model x = （ y^ - a）/b, we can deduce the range of independent variable X from the range of dependent variable Y, But the method is not logical in statistics, The article oppugns the method of traditional statistical control by theory of and furnish two methods to substitute traditional method.
出处
《数理统计与管理》
CSSCI
北大核心
2005年第5期64-67,共4页
Journal of Applied Statistics and Management
关键词
回归模型
统计控制
regression model
statistical control
