Swiss Summer School 2003

Tony Coxon
Multidimensional Scaling

Tony Coxon

Tony Coxon is Emeritus Professor of Sociological Research Methods at the University of Wales, and until recently was Research Professor at the University of Essex. He regularly teaches courses on multidimensional scaling (MDS) at the Essex Summer School in Social Science Data Analysis, at the ICPSR Summer Program in Quantitative Methods , the Master's course in Quantitative Analysis at the Katholieke Universiteit van Brussel . His research interests outside MDS include Classification and Sorting Methods, Social stratification, Sociology of sexualities and integration of qualitative and quantitative information. (see for further information).

Workshop contents and objectives

The aim of the workshop is to show that many sorts of empirical information can be interpreted as "dissimilarity" measures and represented graphically as a map where inter-point distances (or angular separations) reflect those measures as closely as possible. MDS is thus akin to factor analysis, clustering and correspondence analysis and the course will show how these methods are inter-related and how MDS models can provide a framework for analysis of data as different as correlations, tables, categories, measures of association, multiple time and multi-mode data . Emphasis is placed not only the wide range of available models, but also on the acceptability and interpretation of scaling results.

Students will be introduced to using NewMDSX for Windows ( ) and SPSS for the multidimensional analysis of data, and opportunity will exist for "data surgeries" for discussion of appropriate techniques for the analysis of participants' own data.


  1. Borg, I. & P. Groenen (1997) Modern multidimensional Scaling: theory and applications: New York: Springer
  2. Cox, T.F. & M.A.A. Cox (1994) Multidimensional Scaling, London: Chapman and Hall
  3. Coxon, A.P.M. (1982) The Users Guide to multidimensional scaling, London: Heinemann (available online)
  4. Kruskal, J.B. & M. Wish (1978) Multidimensional Scaling , London: Sage QASS no. 11


A working understanding of descriptive statistics and regression is sufficient, but some familiarity with simple geometry and linear algebra will be an advantage.