Is Schwartz's Value Survey an interval scale, and does it really matter?

Abstract

Researchers often assume the numerical ratings approach used to measure values, such as Schwartz's Value Survey (SVS), conforms to an interval scale. Correspondence analysis was used to examine this assumption by analyzing SVS data obtained from four Anglo (Australia, New Zealand, United Kingdom, and United States) and two Asian (South Korea and China) countries. The analysis suggested the SVS did not exhibit the characteristics of an interval scale, with responses across all countries producing larger intervals at the low end of the scale and smaller intervals from the mid to high end of the scale. Further analysis suggested there were significant differences in the traditional SVS means and the means suggested by the correspondence analysis. However, when correlations and Euclidian distances between SVS and correspondence analysis scores were examined, they were very high, suggesting the lack of interval scaling was unlikely to affect the relationships between the SVS value types and other constructs. © The Author(s) 2010.