Translating Tests

In: European psychologist, Band 1, Heft 2, S. 89-99

Abstract

With the increasing interest in cross-cultural research, there is a growing need for standard and validated practices for translating psychological instruments. Developing a psychologically acceptable instrument for another cultural group almost always requires more effort than a literal translation, which all too often is the common practice. The adequacy of translations can be threatened by various sources of bias. Three types of bias are distinguished in this paper: (1) construct bias (related to nonequivalence of constructs across cultural groups), (2) method bias (resulting from instrument administration problems), and (3) item bias (often a result of inadequate translations such as incorrect word choice). Ways in which bias can affect the adequacy of instruments are illustrated and possible remedies are discussed.