A program is described the purpose of which is the assessment of the importance of the independent variables in a multiple regression equation. Five alternate importance measures are provided with primary focus on the statistic proposed by Green, Carroll, and DeSarbo (1978). Input to the program can be in raw data or matrix form and can appear on punched cards or on alternate input media. The program is limited to 20 independent variables but can be easily modified by the user if desired.
Get full access to this article
View all access options for this article.
References
1.
Darlington, R.Multiple regression in psychological research and practice. Psychological Bulletin, 1968, 69, 161-182.
2.
Green, Paul E., Carroll, J. Douglas, and DeSarbo, Wayne S., A new measure of predictor variable importance in multiple regression . Journal of Marketing Research, 1978, 15, 356-360.
3.
Gibson, W. A.Orthogonal predictors: A possible resolution to the Hoffman-Ward controversy. Psychological Reports , 1962, 11, 32-34.
4.
Johnson, Richard M.The minimal transformation to orthonormality. Psychometrika, 1966, 31, 61-66.
5.
IMSL, Inc. IMSL Library (Ver. 7.0, Houston, 1979.
6.
Lutz, J. G.A note on the computation of delta squared. Under review by the Journal of Marketing Research , 1979.
