A Validation Study of the German Complex-Span Tasks and Some General Considerations on Task Translation Procedures in Cognitive Psychology

Abstract

Abstract. Automated complex-span tasks are widely used to assess working-memory capacity and the English versions show good psychometric properties ( Unsworth, Heitz, Schrock, & Engle, 2005 ). However, it is generally an open question whether translated task versions have the same properties as the original versions and whether results obtained with translated tasks can be interpreted equivalently to those obtained with the original tasks. We translated the complex-span tasks and had a sample of German participants perform these tasks as well as a running-memory-span task and a reasoning test. We assessed the reliabilities of the German complex-span tasks and their construct and criterion-related validities. Extrapolating from cross-cultural literature, we also employed a test of measurement invariance to compare the correlational patterns as well as the construct structure between the German sample and a similar North-American sample. Results show that the German complex-span tasks are reliable and valid indicators of working-memory capacity and that they are metrically and functionally equivalent to the original versions. As measurement equivalence is an important but often neglected topic in basic cognitive psychology, we also highlight the general benefits of using equivalence tests when translating cognitive tasks.