Authors: Barbara M. Byrne
Addresses: School of Psychology, University of Ottawa, Canada
Abstract: Scientific inquiry that focuses on multi-group comparisons across national borders can be fraught with problems. Of critical import is empirical evidence that the assessment scale is operating equivalently across groups with respect to perceived item content and dimensionality of the underlying construct. When comparisons are conducted across culture, the process is complex and becomes increasingly so as the number of groups increases and the geo-cultural bases diverge. Structural equation modelling (SEM) provides the most rigorous approach to testing for these equivalencies. The intent of this article is to: a) highlight commonly-held assumptions in testing across national groups; b) outline steps involved in testing these assumptions within the framework of SEM; c) identify three types of test bias effects and exemplify how they can impact cross-national comparisons; d) describe and evaluate translation versus adaptation of a measuring instrument from one language and culture to another language for use in another culture.
Keywords: cross-cultural research; measurement scale equivalence; structural equation modelling; SEM; testing; national groups; test bias effects; measuring instruments; measuring instrument adaptation; measuring instrument translation; language; culture.
International Journal of Business and Globalisation, 2015 Vol.14 No.2, pp.170 - 186
Available online: 11 Feb 2015 *Full-text access for editors Access for subscribers Purchase this article Comment on this article