The Three Society Study

Abstract

In reading the work of Schwartz, and Triandis in the 1990s, I was impressed by the improvements that had occurred in measuring values. It seemed to me that these new methods could be used to answer old questions about the organization of values. Just as in the study of cultural classification systems, techniques of analysis had moved from attempts at feature analysis to the use of multidimensional scaling (Romney 1989); so the study of values had also moved from the intuitive organization of values to the use of multidimensional analysis, especially component analysis, including principal components analysis. I thought that if a very large and varied set of value items could be constructed, ratings of values by respondents from different cultures could be subjected to component analysis, including principal components analysis, and the dimensions of values uncovered. None of the quantitative value studies had used really large numbers of value items. Using 200 or 300 value items, I thought, would make it possible to uncover both universal and culturally specific dimensions for different cultures. Rokeach and Schwartz had not made use of principal components analysis. The result might be the discovery of the kind of dimensions that the Kluckhohns had been searching for.