Some Thoughts on Variable-Selection in Multiple Regression

Article

Mandel, John   (1989, ASQC)   National Bureau of Standards

Journal of Quality Technology    Vol. 21    No. 1
QICID: 5635    January 1989    pp. 2-6
List $10.00
Member $5.00

FOR A LIMITED TIME, ACCESS TO THIS CONTENT IS FREE!
You will need to be signed in.
New to ASQ? Register here.

Article Abstract

The practice of variable-selection in multiple linear regression is examined both logically and statistically in terms of the major objectives of regression analysis. It is argued that in addition to goodness-of-fit considerations, due attention should be given to determining the domain, in the space of regressors, in which the regression equation can be used as a prediction device. Such a domain has been called the "effective prediction domain" (EPD). It is shown that failure to do so can result in serious misuse of the regression equation, especially when regressors have been omitted.

Keywords

Process capability (Cp),Goodness of fit,Multiple regression,Prediction,Selection,Collinearity,Regression


Browse QIC Articles Chronologically:     Previous Article     Next Article

New Search

Featured advertisers





ASQ is a global community of people passionate about quality, who use the tools, their ideas and expertise to make our world work better. ASQ: The Global Voice of Quality.