This article gives an overview of Facet Theory, a systematic approach to facilitating theory construction, research design, and data analysis for complex studies, that is particularly appropriate to the behavioral and social sciences. Facet Theory is based on (1) a definitional framework for a universe of observations in the area of study; (2) empirical structures of observations within this framework; (3) a search for correspondence between the definitional system and aspects of the empirical structure for the observations. The development of Facet Theory and Facet Design is reviewed from early scale analysis and the Guttman Scale, leading to the concepts of “mapping sentence,” “universe of content,” “common range,” “content facets,” and nonmetric multidimensional methods of data analysis. In Facet Theory, the definition of the behavioral domain provides a rationale for hypothesizing structural relationships among variables employed in a study. Examples are presented from various areas of research (intelligence, infant development, animal behavior, etc.) to illustrate the methods and results of structural analysis with Smallest Space Analysis (SSA), Multidimensional Scalogram Analysis (MSA), and Partial Order Scalogram Analysis (POSA). The “radex” and “cylindrex” of intelligence tests are shown to be outstanding examples of predicted spatial configurations that have demonstrated the ubiquitous emergence of the same empirical structures in different studies. Further examples are given from studies of spatial abilities, infant development, animal behavior, and others. The use of Facet Theory, with careful construction of theory and design, is shown to provide new insights into existing data; it allows for the diagnosis and discrimination of behavioral traits and makes the generalizability and replication of findings possible, which in turn makes possible the discovery of lawfulness. Achievements, issues, and future challenges of Facet Theory are discussed.
Reconstructing Point Sets from Distance Distributions