A general nonmetric technique for finding the smallest coordinate space for a configuration of points

LetA1,A2, ...,An be anyn objects, such as variables, categories, people, social groups, ideas, physical objects, or any other. The empirical data to be analyzed are coefficients of similarity or distance within pairs (Ai,Ai), such as correlation coefficients, conditional probabilities or likelihoods, psychological choice or confusion, etc. It is desired to represent these data parsimoniously in a coordinate space, by calculatingm coordinates {xia} for eachAi for a semi-metricd of preassigned formdij =d(|xi1 -xj1|, |xi2 -xj2|, ..., |xim -xjm|). The dimensionalitym is sought to be as small as possible, yet satisfy the monotonicity condition thatdij <dkl whenever the observed data indicate thatAi is “closer” toAj thanAk is toAl. Minkowski and Euclidean spaces are special metric examples ofd. A general coefficient of monotonicityμ is defined, whose maximization is equivalent to optimal satisfaction of the monotonicity condition, and which allows various options both for treatment of ties and for weighting error-of-fit. A general rationale for algorithm construction is derived for maximizing μ by gradient-guided iterations; this provides a unified mathematical solution to the basic operational problems of norming the gradient to assure proper convergence, of trading between speed and robustness against undesired stationary values, and of a rational first approximation. Distinction is made between single-phase (quadratic) and two-phase (bilinear) strategies for algorithm construction, and between “hard-squeeze” and “soft-squeeze” tactics within these strategies. Special reference is made to the rank-image and related transformational principles, as executed by current Guttman-Lingoes families of computer programs.

[1]  L. Guttman,et al.  The Quantification of a class of attributes : A theory and method of scale construction , 1941 .

[2]  Louis Guttman,et al.  An Approach for Quantifying Paired Comparisons and Rank Order , 1946 .

[3]  Louis Guttman,et al.  A generalized simplex for factor analysis , 1955 .

[4]  AN ADDITIVE METRIC FROM ALL THE PRINCIPAL COMPONENTS OF A PERFECT SCALE1 , 1955 .

[5]  Louis Guttman,et al.  Introduction to facet design and analysis , 1959 .

[6]  Louis Guttman,et al.  A STRUCTURAL THEORY FOR INTERGROUP BELIEFS AND ACTION , 1959 .

[7]  C. Coombs A theory of data. , 1965, Psychology Review.

[8]  R. Shepard The analysis of proximities: Multidimensional scaling with an unknown distance function. II , 1962 .

[9]  S. Messick,et al.  Scaling and Measurement Theory , 1963 .

[10]  S. Messick,et al.  Chapter VII: Scaling and Measurement Theory , 1963 .

[11]  J. Kruskal Nonmetric multidimensional scaling: A numerical method , 1964 .

[12]  J. Kruskal Multidimensional scaling by optimizing goodness of fit to a nonmetric hypothesis , 1964 .

[13]  W. Torgerson,et al.  Multidimensional scaling of similarity , 1965, Psychometrika.

[14]  T. Mori Structure of motivations for becoming a teacher. , 1965, Journal of educational psychology.

[15]  Louis Guttman,et al.  The Relative Associational Contiguity of Occupations in an Urban Setting , 1966 .

[16]  L. Guttman,et al.  Cross-ethnic variation in dental, sensory and perceptual traits: a nonmetric multibivariate derivation of distances for ethnic groups and traits. , 1967, American journal of physical anthropology.

[17]  L Guttman,et al.  Nonmetric Factor Analysis: A Rank Reducing Alternative To Linear Factor Analysis. , 1967, Multivariate behavioral research.

[18]  I. M. Schlesinger,et al.  Systematic Construction of Distractors for Ability and Achievement Test Items , 1967 .

[19]  L Guttman,et al.  THE DEVELOPMENT OF NONMETRIC SPACE ANALYSIS: A LETTER TO PROFESSOR JOHN ROSS. , 1967, Multivariate behavioral research.

[20]  M Wish,et al.  A Model for the Perception of Morse Code-Like Signals1 , 1967, Human factors.

[21]  Forrest W. Young Computer program abstracts , 1968 .

[22]  J. Lingoes The Multivariate Analysis Of Qualitative Data , 1968 .

[23]  Louis Guttman,et al.  Smallest space analysis of intelligence and achievement tests. , 1969 .