Because of several problems which are inherent in the use of traditional algorithms for selecting a subset of predictor variables for a multiple regression equation, an alternate algorithm and computer program are described which use expected cross-validated correlation rather than multiple correlation as the criterion for selecting variables. All usual multiple regression information, as well as regression weights on the subset of variables chosen and the cross-validation correlation to be expected on future applications of these weights, is output. Documentation is provided.
Get full access to this article
View all access options for this article.
References
1.
Browne, M.W.Predictive validity of a linear regression equation. British Journal of Mathematical and Statistical Psychology, 1975, 28, 79-87.
2.
Cattin, P.Note on the estimation of the squared cross-validated multiple correlation of a regression model. Psychological Bulletin, 1980, 87, 63-85. (a)
3.
Cattin, P.Estimation of the prediction power of a regression model. Journal of Applied Psychology, 1980, 65, 407-414. (b)
4.
Draper, N.R. and Smith, H.Applied regression analysis. New York: Wiley, 1966.
5.
Drasgow, F., Dorans, N.J., and Tucker, L.R.Estimators of the squared cross-validity coefficient: A Monte Carlo investigation . Applied Psychological Measurement, 1979 , 3, 387-399.
6.
Olkin, I. and Pratt, J.W.Unbiased estimation of certain correlation coefficients. Annals of Mathematical Statistics, 1958, 29, 201-211.
7.
Rozeboom, W.W.Estimation of cross-validated multiple correlation: A clarification . Psychological Bulletin, 1978, 85, 1348-1351.
8.
Schmitt, N., Coyle, B.W., and Rauschenberger, J.A Monte Carlo evaluation of three formula estimates of cross-validated multiple correlation. Psychological Bulletin, 1977, 84, 751-758.
9.
Tatsuoka, M.M.Multivariate analysis in educational research. In F. N. Kerlinger (Ed.), Review of research in education (Vol. 1). Itasca, Ill.: Peacock, 1973.
