One statistical technique may not be enough: comparing measurement of psychological attributes across countries

Abstract

The aim of this study was to compare the impact of the statistical techniques on measurement outcomes of Collectivism and Individualism scales across two countries, New Zealand (NZ) and China, and five ethnic groups. Attributes of Collectivism and Individualism (AICS) were compared across undergraduate students from Mainland China and New Zealand. Two statistical techniques (t-test and k-mean cluster analysis) were used to compare the samples and identify differences. The mean scores on both Collectivism and Individualism for the Chinese sample were lower. These results suggest that mean scores (e.g., t-test) of psychological measures do not provide enough information for international comparison and are therefore more susceptible to response biases than population classification-based analysis (e.g., cluster analysis). Implications for other psychological measures are discussed.