Candelinc: A General Approach to Multidimensional Analysis of Many-Way Arrays with Linear Constraints on Parameters | Psychometrika | Cambridge Core (original) (raw)
Abstract
Very general multilinear models, called CANDELINC, and a practical least-squares fitting procedure, also called CANDELINC, are described for data consisting of a many-way array. The models incorporate the possibility of general linear constraints, which turn out to have substantial practical value in some applications, by permitting better prediction and understanding. Description of the model, and proof of a theorem which greatly simplifies the least-squares fitting process, is given first for the case involving two-way data and a bilinear model. Model and proof are then extended to the case of _N_-way data and an _N_-linear model for general N. The case N = 3 covers many significant applications. Two applications are described: one of two-way CANDELINC, and the other of CANDELINC used as a constrained version of INDSCAL. Possible additional applications are discussed.
References
Carroll, J. D., Green, P. E., & Carmone, F. J. CANDELINC: A new method for multidimensional analysis with constrained solutions. Paper presented at the meeting of the International Congress of Psychology, Paris, France, July 1976.Google Scholar
Carroll, J. D., & Pruzansky, S. MULTILINC: Multiway CANDELINC. Paper presented at the meeting of the American Psychological Association, San Francisco, August 1977.Google Scholar
Carroll, J. D. & Pruzansky, S. Use of LINCINDS as a rational starting configuration for INDSCAL. Unpublished memorandum, Bell Telephone Laboratories, 1979.Google Scholar
Cohen, H. S. Three-mode rotation to approximate INDSCAL structure. Paper presented at the annual meeting of the Psychometric Society, Palo Alto, California, March, 1974.Google Scholar
DeLeeuw, J. & Heiser, W. Multidimensional scaling with restrictions on the configuration. Proceedings of the Fifth International Symposium on Multivariate Analysis, Pittsburgh, June 1978.Google Scholar
Harshman, R. A. Foundations of the PARAFAC procedure; Models and conditions for an “explanatory” multi-modal factor analysis. UCLA Working Papers in Phonetics, 1970, 16, 1–84.Google Scholar
Harshman, R. A. PARAFAC-2; mathematical and technical notes. UCLA Working Papers in Phonetics, 1972, 22, 30–44.Google Scholar
Kruskal, J. B. & Carmone, F. How to use M-D-SCAL (Version 5M) and other useful information. Unpublished memorandum, Bell Telephone Laboratories, 1969.Google Scholar
Kruskal, J. B., Young, F. W., & Seery, J. B. How to use KYST, a very flexible program to do multidimensional scaling and unfolding. Unpublished memorandum, Bell Telephone Laboratories, 1973.Google Scholar
Noma, E., & Johnson, J. Constraining nonmetric multidimensional scaling configurations, 1977, Ann Arbor: The University of Michigan, Human Performance Center.Google Scholar
Tucker, L. R. & MacCallum, R. C. Personal communication, 1974.Google Scholar
Bentler, P. M., & Weeks, D. G. Restricted multidimensional scaling models. Journal of Mathematical Psychology, 1978, 17, 138–151.CrossRefGoogle Scholar
Bloxom, B. Constrained multidimensional scaling in N spaces. Psychometrika, 1978, 43, 397–408.CrossRefGoogle Scholar
Borg, I. & Lingoes, J. C. A model and algorithm for multidimensional scaling with external constraints on the distances. Psychometrika, 1980, 45, 25–38.CrossRefGoogle Scholar
Carroll, J. D. Individual differences and multidimensional scaling. In Shepard, R. N., Romney, A. K. & Nerlove, S. (Eds.), Multidimensional scaling theory and applications in the behavioral sciences (Vol. I), 1972, New York: Academic Press.Google Scholar
Carroll, J. D., & Chang, J. J. Analysis of individual differences in multidimensional scaling via an N-way generalization of “Eckart-Young” decomposition. Psychometrika, 1970, 35, 283–320.CrossRefGoogle Scholar
Carroll, J. D., & Wish, M. Multidimensional perceptual models and measurement methods. In Carterette, E. C. & Friedman, M. P. (Eds.), Handbook of perception. (Vol. II) Psychophysical judgement and measurement. New York: Academic Press, Inc.. 1974, 391–447.CrossRefGoogle Scholar
Eckart, C., & Young, G. The approximation of one matrix by another of lower rank. Psychometrika, 1936, 1, 211–218.CrossRefGoogle Scholar
Green, P. E., Carroll, J. D., & Carmone, F. J. Superordinate factorial designs in the analysis of consumer judgments. Journal of Business Research, 1976, 4, 281–295.CrossRefGoogle Scholar
Horan, C. B. Multidimensional scaling: Combining observations when individuals have different perceptual structures. Psychometrika, 1969, 34, 139–165.CrossRefGoogle Scholar
Johnson, R. M. Minimal transformation to orthonormality. Psychometrika, 1966, 31, 61–66.CrossRefGoogle Scholar
MacCallum, R. C. Effects of conditionality on INDSCAL and ALSCAL weights. Psychometrika, 1977, 42, 297–305.CrossRefGoogle Scholar
McGee, V. E. Multidimensional scaling of N sets of similarity measures: A nonmetric individual differences approach. Multivariate Behavioral Research, 1968, 3, 233–248.CrossRefGoogle Scholar
Torgerson, W. S. Theory and Methods of Scaling, 1958, New York: Wiley.Google Scholar
Tucker, L. The extension of factor analysis to three-dimensional matrices. In Frederiksen, N. & Gulliksen, H. (Eds.), Contributions to mathematical psychology. New York: Holt, Rinehart & Winston. 1964, 109–127.Google Scholar
Tucker, L. R. Relations between multidimensional scaling and three-mode factor analysis. Psychometrika, 1972, 37, 3–27.CrossRefGoogle Scholar
Wish, M. Subjects' expectations about their own interpersonal communication: A multidimensional approach. Personality and Social Psychology Bulletin, 1975, 1, 501–504.CrossRefGoogle Scholar
Wish, M. & Kaplan, S. J. Toward an implicit theory of interpersonal communication. Sociometry, 1977, 40, 234–246.CrossRefGoogle Scholar
Wold, H. Estimation of principal components and related models by iterative least squares. In Krishnaiah, P. R. (Eds.), Multivariate Analysis. New York: Academic Press. 1966, 391–420.Google Scholar