Quality variations of psychometrics for cross-cultural research

Abstract

The study examined the quality variations of psychometrics for cross-cultural research by considering structural validity, discrimination, and reliability. Participants consisted of 450 undergraduate students from Thailand, Indonesia, and Australia by using random sampling process. The instruments were driving aggression evaluative forms which contain 0.85 alpha coefficients. For statistics, researchers used mean, standard deviation, correlation coefficient among point-per-questions and total score, and Cronbach’s alpha. The result showed that the structural validity of the original evaluation form which was translated once from Thai to English and another one which was translated twice including Thai to English and English to Indonesian had similar total components and cumulative percentiles. Discrimination and correlation among point-per-questions met the criteria without differences. Every form had the level of reliability from Cronbach’s alpha analysis greater than the criteria.