A strategy for reordering the hierarchical tree structure is presented. While the order of terminal nodes of Johnson's (1967) procedure is arbitrary, depending simply upon the order of input data, the present procedure will rearrange every triad of nodes under a common least upper node at all levels so that the node in the middle is always non-arbitrarily closest to the anchored node. Solutions for data from Miller and Nicely (1955) by Johnson and two rotational schemes are discussed.
Get full access to this article
View all access options for this article.
References
1.
Gruvaeus, G. and Wainer, H.Two additions to hierarchical cluster analysis. British Journal of Mathematical and Statistical Psychology, 1972, 25, 200-206.
2.
Halff, H. M.Graphical evaluation of hierarchical clustering schemes, Technical Report No. 1, Laboratory for Cognitive Studies in Education, University of Illinois, 1975.
3.
Harman, H. H.Modern factor analysis. Chicago: University of Chicago Press, 1967.
4.
Johnson, S. C.Hierarchical clustering schemes . Psychometrika, 1967, 32, 241-254.
5.
Miller, C. A. and Nicely, P. E.An analysis of perceptual confusion among some English consonants. Journal of the Acoustical Society of America , 1955, 27, 338-352.
6.
Tucker, L. R. and Messick, S.An individual differences model for multidimensional scaling. Psychometrika. 1963, 28, 333-367.
7.
Tzeng, O. C. S.On the construct validity of multidimensional scaling through the INDSCAL model. Paper presented at the Joint Meeting of the Psychometric Society and the Classification Society. Iowa, April, 1975.
8.
Tzeng, O. C. S., May, W. H., and Osgood, C. E.A context-free culture/concept contingency analysis of attribution of affect in 23 human communities Center for Comparative Psycholinguistics, University of Illinois at Urbana-Champaign, 1973.
