Scaling versus Content Analysis: Interpreting Word Association Data from Americans and Koreans

Abstract

Korean and American cultures are compared with respect to the way in which semantic domains are organized. With similarity scores based on the results of a modified word association test, words from different semantic domains are analyzed by means of the Kruskal-Shepard multidimensional scaling technique and by content analysis. The results from the two techniques are then compared. Multidimensional scaling proves to be an effective technique for mapping a large set of interrelationships into a simple spatial configuration, but it fails to yield results clearly indicative of the differences between Korean and American culture. Content analysis, while less integrative as a technique of data analysis, yields results which appear to be more representative of differences between Korean and American culture.