Hostname: page-component-745bb68f8f-hvd4g Total loading time: 0 Render date: 2025-01-07T19:25:40.775Z Has data issue: false hasContentIssue false
  • English
  • Français

Constrained Canonical Correlation

Published online by Cambridge University Press:  01 January 2025

Wayne S. DeSarbo ,
Robert E. Hausman ,
Shen Lin  and
Wesley Thompson
Wayne S. DeSarbo*
Affiliation:
Bell Laboratories
Robert E. Hausman
Affiliation:
Bell Laboratories
Shen Lin
Affiliation:
Bell Laboratories
Wesley Thompson
Affiliation:
Bell Laboratories
*
Requests for reprints should be addressed to: Wayne S. DeSarbo, Bell Labs., Room 2C-479, 600 Mountain Ave., Murray Hill, N.J. 07974.
Get access Rights & Permissions [Opens in a new window]

Abstract

This paper explores some of the problems associated with traditional canonical correlation. A response surface methodology is developed to examine the stability of the derived linear functions, where one wishes to investigate how much the coefficients can change and still be in an ɛ-neighborhood of the globally optimum canonical correlation value. In addition, a discrete (or constrained) canonical correlation method is formulated where the derived coefficients of these linear functions are constrained to be in some small set, e.g., {1, 0, −1}, to aid in the interpretation of the results. An example concerning the psychographic responses of Wharton MBA students of the University of Pennsylvania regarding driving preferences and life-style considerations is provided.

Keywords

canonical correlationconstrained multivariate analysisresponse surface analysis
Type
Original Paper
Information
Psychometrika , Volume 47 , Issue 4 , December 1982 , pp. 489 - 516
Copyright
Copyright © 1982 The Psychometric Society

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

Footnotes

Wayne S. DeSarbo, Robert Jausman, Shen Lin, and Wesley Thompson are all Members of Technical Staff at Bell Laboratories. We wish to express our gratitude to the editor and reviewers of this paper for their insightful remarks.

References

Reference Notes

Hausman, R., Constrained multivariate analysis, Unpublished Memorandum, Bell Laboratories, Holmdel, N.J., 1980a.Google Scholar
Hausman, R., Constrained principal components, Unpublished Memorandum, Bell Laboratories, Holmdel, N.J., 1980b.Google Scholar
Kettenring, J. R. Canonical analysis, 1979, Murray Hill, N.J.: Bell Laboratories.Google Scholar

References

Anderson, T. W. Introduction to Multivariate Statistical Analysis, 1958, New York: John Wiley & Sons.Google Scholar
Bard, Y. Nonlinear Parameter Estimation, 1974, New York: Academic Press.Google Scholar
Bartlett, M. S. On the theory of statistical regression. Proceedings of the Royal Society of Edinburgh, 1933, 53, 260–283.CrossRefGoogle Scholar
Bentler, P. M., & Woodwards, J. A., Regression on linear composites: Statistical theory and applications. Multivariate Behavioral Research Monographs, 1979, No. 79-1.Google Scholar
Bibby, J., & Toutenberg, H. Prediction and improved estimation in linear models, 1977, New York: Wiley.Google Scholar
Bibby, J. Some Effects of Rounding Optimal Estimates, 1980, England: The Open University.Google Scholar
Chambers, J. Computational Methods for Data Analysis, 1977, New York: John Wiley & Sons.Google Scholar
Cliff, N. & Krus, D. J. Interpretation of canonical analysis: rotated vs. unrotated solutions. Psychometrika, 1976, 4, 35–42.CrossRefGoogle Scholar
Cooley, W. W., Canonical correlation, Paper presented at the 74-th Annual Convention, American Psychological Association (mimeograph), 1965.Google Scholar
Cooley, W. W. & Lohnes, P. R. Multivariate Procedures for the Behavioral Sciences, 1962, New York: John Wiley & Sons.Google Scholar
Darlington, R. B. & Bishop, C. H. Increasing test validity by considering interitem correlations. Journal of Applied Psychology, 1966, 50, 322–330.CrossRefGoogle ScholarPubMed
Darlington, R. B. Reduced-variance regression. Psychological Bulletin, 1978, 85, 1238–1255.CrossRefGoogle ScholarPubMed
Dempster, A. P. Estimation in multivariate analysis. In Krishnaiah, P. R. (Eds.), Multivariate Analysis. New York: Academic Press. 1966, 315–334.Google Scholar
Garfinkel, R. S. and Nemhauser, G. L. Integer Programming, 1972, New York: John Wiley & Sons.Google Scholar
Gnanadesikan, R. Methods for Statistical Data Analysis of Multivariate Observations, 1977, New York: John Wiley & Sons.Google Scholar
Green, B. R. Jr. Parameter sensitivity in multivariate methods. Multivariate Behavioral Research, 1977, 12, 263–287.CrossRefGoogle ScholarPubMed
Green, P. E., Carroll, J. D., & DeSarbo, W. S. An extended application of the δ2 measure of predictor-variable importance, 1979, Chicago: A.M.A..Google Scholar
Hotelling, H. Relations between two sets of variates. Biometrika, 1936, 28, 321–377.CrossRefGoogle Scholar
Kendall, M. G., and Stuart, A. S. The Advanced Theory of Statistics, Vol. 3, 1966, New York: Hafner Publishing Co..Google Scholar
Kendall, M. G. A Course in Multivariate Analysis, 1968, New York: Hafner Publishing Co..Google Scholar
Levine, M. S. Canonical Analysis and Factor Comparisons, 1977, Beverly Hills: Sage Publications.CrossRefGoogle Scholar
Lin, S. and Kernighan, M. An effective heuristic algorithm for the traveling salesman problem. Operations Research, 1973, 21, 498–516.CrossRefGoogle Scholar
Meredith, W. Canonical correlations with fallible data. Psychometrika, 1964, 29, 55–65.CrossRefGoogle Scholar
Pruzek, R. M., & Frederick, R. C. Weighting predictors in linear models: Alternatives to least squares and limitations of equal weights. Psychological Bulletin, 1978, 85, 254–266.CrossRefGoogle Scholar
Reingold, E. M., Nievergelt, J., & Deo, N. Combinatorial Algorithms, 1977, Englewood Cliffs, New Jersey: Prentice Hall.Google Scholar
Roy, S. N. Some Aspects of Multivariate Analysis, 1957, New York: John Wiley and Sons, Inc..Google Scholar
Scheffé, H. The Analysis of Variance, 1959, New York: John Wiley and Sons, Inc..Google Scholar
Thorndike, R. M., Weiss, D. J., & Davis, R. V. Canonical correlation of vocational interests and vocational needs. Journal of Counseling Psychology, 1968, 15, 101–106.CrossRefGoogle Scholar
Thorndike, R. M. &, Weiss, D. J. A Study of the stability of canonical correlations and canonical components. Educational and Psychological Measurement, 1973, 33, 123–134.CrossRefGoogle Scholar
Wainer, H. Estimating coefficients in linear models: It don't make no nevermind. Psychological Bulletin, 1976, 83, 213–217.CrossRefGoogle Scholar
Wainer, H. On the sensitivity of regression and regressors. Psychological Bulletin, 1978, 85, 267–273.CrossRefGoogle Scholar
Weinberg, S. L. & Darlington, R. B. Canonical analysis when number of variables is large relative to sample size. Journal of Educational Statistics, 1976, 1, 313–332.CrossRefGoogle Scholar
Winer, B. J. Statistics and data analysis: Trading bias for reduced mean square error. Annual Review of Psychology, 1978, 29, 647–6811.CrossRefGoogle Scholar
Vinod, H. D. Canonical ridge and econometrics of joint production. Journal of Econometrics, 1977, 4, 147–166.CrossRefGoogle Scholar