7 Discussion

 

In the present study, many useful properties of the converging sequence of iteratively computed correlation matrices given a proximity matrix have been introduced. Structures in and mimic the effects in dimension reduction techniques such as factor analysis and multidimensional scaling. Near stationary iterations with the sorted colored maps can be employed to identify structural (clustering) information embedded in the data. A rank-two iteration finds the Robinson seriation in the proximity matrix while the converged rank-one structure splits a proximity matrix into two sets with the divisive clustering tree and the rank-one tree seriation. A non-rank-one converged pattern matrix can also be used to study the symmetry pattern exists in the proximity matrix. The whole converging process projected through the eigenvector plot with ellipse is a powerful and dynamic visualization environment for many faces of high-dimensional statistical data analysis.

 

The original purpose of this study was to investigate the general behavior of patient-clusters on symptom-groups for the psychosis disorder data set. Instead of using many of the available multivariate analysis methods, we have come up with the generalized association plots (GAP) for information visualization.

 

We have partially accomplished the goal. Through careful examination of the double sorted raw data matrix map with the sufficient graphs and the sequence of correlation maps for both the symptoms and the patients, we thoroughly understand the symptom-groups and the patient-clusters in detail. However, this is only the beginning of our effort for understanding the whole process of the development of psychosis disorder disease. The fifty SAPS and SANS symptoms used in this study is only part of the many rating scales in the MPGRP project. The complete data base comes with different rating scales (nominal, ordinal, and continuous) at different time points with biological background information of each patient in the study. It is an extremely difficult challenge to develop multiphase longitudinal and categorical versions of GAP to help in understanding this kind of large-scale study in a more thorough manner.

 

References

  • Andreasen, N. C. (1983), "The Scale for the Assessment of Negative Symptoms (SANS)," University of Iowa, Iowa City, IA.

  • Andreasen, N. C. (1984), "The Scale for the Assessment of Positive Symptoms (SANS), "University of Iowa, Iowa City, IA.

  • Andreasen, N. C., Arndt, S., Alliger, R., Miller, D., Flaum, M. (1995), "Symptoms of schizophrenia: methods, meanings, and mechanisms." Archive of General Psychiatry, 52, 341-351.

  • Breiger, R. L., Boorman, S. A. and Arabie, P (1975), "An Algorithm for Clustering Relational Data with Applications to Social Network Analysis and Comparison with Multidimensional Scaling," J. of Math. Psych., 12, 328-383.

  • Caraux, G. (1984), "Rearrangement and Visual Representation of Matrices with Numerical Data: An Iterative Algorithm (French)," Rev. Statist. Appl., 32, 5-23.

  • Chen, C, H., and Chen, J. A. (2000), "Interactive Diagnostic Plots for Multidimensional Scaling with Applications in Psychosis Disorder Data Analysis," to appear in Statistica Sinica, 10.

  • Fisher, R. A. (1936), "The Use of Multiple Measurements in Taxonomic Problems," Ann. Eugen., 7, 179-188.

  • Gale, N., Halperin, C. W., and Costanzo, C. M. (1984), "Unclassed Matrix Shading and Optimal Ordering in Hierarchical Cluster Analysis," Journal of Classification, 1, 75-92.

  • Hubert, L. (1976), "Seriation Using Asymmetric Proximity Measures," British J. Math. Statist. Psych., 29, 32-52.

  • Kruskal, B. J. (1977), "A Theorem about CONCOR," unpublished manuscript.

  • Lin, A. S., Chen, C. H., Hwu H. G., Lin H. N., and Chen J. A. (1998), "Psychopathological Dimensions in Schizophrenia: A Correlational Approach to Items of the SANS and SAPS," Psychiatry Research, 77, 121-130. (SCI)

  • Marcotorchino, F. (1991), "Seriation Problems: An Overview," Applied Stochastic Models and Data Analysis, 7, 139-151.

  • McFarlane, M., and Young, F. W. (1994), "Graphical Sensitivity Analysis for Multidimensional Scaling," Journal of Computational and Graphical Statistics, 3, 23-33.

  • McQuitty, L. L. (1968), "Multiple Clusters, Types, and Dimensions from Iterative Intercolumnar Correlational Analysis," Multivariate Behavioral Research, 3, 465-477.

  • Minas, I. H., Stuart, G. W., Klimidis, S., Jackson, H. J., Singh, B. S. and Copolov, D. L. (1992), "Positive and negative symptoms in the psychoses: Multidimensional scaling of SAPS and SANS items." Schizophrenic Research, 8, 143-156.

  • Minnotte, M., and West, R. W. (1998), "The Data Image: A tool for exploring high dimensional data sets," 1998 Proceedings of the Section on Statistical Graphics, American Statistical Association.

  • Murdoch, D. J., and Chow, E. D. (1996), "A Graphical Display of Large Correlation Matrices," Statistical Computing, 50, 178-180.

  • Robinson, W. S. (1951), "A Method for Chronologically Ordering Archaeological Deposits," American Antiquity, 16, 293-301.

  • Sokal, R. R., and Sneath, P. H. (1963), Principles of Numerical Taxonomy, San Francisco: Freeman.

  • Streng, R. (1991) "Classification and Seriation by Iterative Reordering of a Data Matrix," In Classification, Data Analysis, and Knowledge Organization: Models and Methods with Applications, 121-130, edited by Bock, H. H., and Ihm, P. New York: Springer-Verlag.

  • Stuart, G. W., Malone, V., Currie, J., Klimidis, S. and Minas, I. H. (1995), "Positive and negative symptoms in neuroleptic-free psychotic inpatients." Schizophrenic Research, 16, 175-188.

  • Takane, Y., Young, F. W., and de Leeuw, J. (1977), "Nonmetric Individual Differences Multidimensional Scaling: An Alternating Least Squares Method with Optimal Scaling Features," Psychometrika, 42, 7-67.

  • Tierney, L. (1990), Lisp-Stat: An Object-Oriented Environment for Statistical Computing and Dynamic Graphics, New York: John Wiley.

 

[Prev] [Context] [Next]