Hostname: page-component-7b7567c8c-mnklg Total loading time: 0 Render date: 2026-11-12T09:43:55.455Z Has data issue: false hasContentIssue false
  • English
  • Français

The Optimality of the Centroid Method

Published online by Cambridge University Press:  01 January 2025

Vartan Choulakian
Vartan Choulakian*
Affiliation:
Départment de Mathématiques et de Statistique, Université de Moncton
*
Requests for reprints should be sent to Vartan Choulakian, Départment de Mathématiques et de Statistique, Université de Moncton, and Moncton, N.B., CANADA E1A 3E9. E-Mail: choulav@umoncton.ca
Get accessRights & Permissions [Opens in a new window]

Abstract

The aim of this note is to show that the centroid method has two optimality properties. It yields loadings with the highest sum of absolute values, even in absence of the constraint that the squared component weights be equal. In addition, it yields scores with maximum variance, subject to the constraint that none of the squared component weights be larger than 1.

Keywords

principal component analysis

Information

Type
Article
Information
Psychometrika , Volume 68 , Issue 3 , September 2003, pp. 473 - 475
Copyright
Copyright © 2003 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.)

Article purchase

Temporarily unavailable

Footnotes

This research is financed by NSERC of Canada. The author is grateful to Michel Tenenhaus for pointing the similarity of the procedures in the centroid method and Q-mode PCA in L1. The author also thanks the editor and associate editor for providing shorter proofs of the theorems, along with the referees for their helpful comments.

References

Anderson, T.W. (1984). An introduction to multivariate statistical analysis 2nd ed., New York, NY: John Wiley & Sons.Google Scholar
Benayade, M., Fichet, B. (1994). Algorithms for a geometrical PCA in L1-norm. In Diday, E., Lechevallier, Y., Schader, M., Bertrand, P., Burtschy, B. (Eds.), New approaches in classification and data analysis (pp. 75–84). Berlin, Germany: Springer.10.1007/978-3-642-51175-2_7CrossRefGoogle Scholar
Bentler, P.M. (1986). Structural modeling andPsychometrika: An historical perspective on growth and achievements. Psychometrika, 51, 35–51.10.1007/BF02293997CrossRefGoogle Scholar
Burt, C. (1917). The distribution and relations of educational abilities. London, U.K.: P.S. King & Son.Google Scholar
Choulakian, V. (2001). Robust Q-mode principal component analysis in L1. Computational Statistics and Data Analysis, 37, 135–150.10.1016/S0167-9473(01)00005-6CrossRefGoogle Scholar
Galpin, J.S., Hawkins, D.M. (1987). L1 estimation of a covariance matrix. Computational Statistics and Data Analysis, 5, 305–319.CrossRefGoogle Scholar
Harman, H.H. (1967). Modern factor analysis. Chicago, IL: The University of Chicago Press.Google Scholar
Heiser, W.J. (1987). Correspondence analysis with least absolute residuals. Computational Statistics and Data Analysis, 5, 337–356.10.1016/0167-9473(87)90057-0CrossRefGoogle Scholar
Holley, J.W. (1947). A note on the reflection of signs in the extraction of centroid factors. Psychometrika, 12, 263–265.10.1007/BF02288940CrossRefGoogle ScholarPubMed
Horst, P. (1965). Factor analysis of data matrices. New York, NY: Holt, Rinehart and Winston.Google Scholar
Lawley, D.N., Maxwell, A.E. (1963). Factor analysis as a statistical method. London, U.K.: Butterworths.Google Scholar
Nunnally, J.C. (1978). Psychometric theory. New York, NY: McGraw Hill.Google Scholar
Thurstone, L.L. (1931). Multiple factor analysis. Psychological Review, 38, 406–427.10.1037/h0069792CrossRefGoogle Scholar