A computational simplification suggested by Allen (1971) for an index of multiple regression cross-validation proffered by Drehmer and Morris (1981) was introduced and compared to the Drehmer-Morris algorithm. Results showed the Allen algorithm to be much faster. Thus unnecessary computational expense was avoided and analyses could be undertaken that would otherwise probably not be possible. A computer program to calculate the cross-validation index by Allen's algorithm is described.
Get full access to this article
View all access options for this article.
References
1.
Allen, D. M. (1971). The prediction sum of squares as a criterion for selecting predictor variables. Technical Report 23, Dept. of Statistics, University of Kentucky.
2.
Draper, N. R. and Smith, H. (1981). Applied regression analysis. New York: Wiley.
3.
Drehmer, D. E. and Morris, G. W. (1981). Cross-validation with small samples: An algorithm for computing Gollob's estimator. Educational and Psychological Measurement, 41, 195-200.
4.
Gollob, H. F. (1967). Cross-validation using samples of size one. Paper presented at the American Psychological Association meeting in Washington, D. C., September.
5.
Huberty, C. J. and Mourad, S. A. (1980). Estimation in multiple correlation/prediction. Educational and Psychological Measurement, 40, 101-112.
6.
Lachenbruch, P. A. and Mickey, M. R. (1968). Estimation of error rates in discriminant analysis. Technometrics, 10, 1-11.
7.
Mosteller, F. and Tukey, W. J. (1968). Data analysis, including statistics. In G. Lindzey and Aronson (Eds.), The handbook of social psychology (2nd ed.). Reading, Mass.: Addison-Wesley.
8.
Stone, M. (1974). Cross-validatory choice and assessment of statistical predictions. Journal of the Royal Statistical Society, Series B., 36, 111-147.
9.
Tatsuoka, M. M. (1971). Multivariate analysis. New York: Wiley.
