In this study, a German and Danish language version of the Attitudes towards and Relationship to Statistics - Revised (original Danish title: “Holdninger og Forhold til Statistik”, HFS-R) instrument was administered to samples of psychology students in Denmark and the German-speaking part of Switzerland. The data were analysed by means of the Rasch model and the extended graphical log-linear Rasch models to assess Rasch model violations. In particular, it was investigated whether items showed differential item functioning (DIF) between the Danish and Swiss students. The results show that a comparison between the students in the different countries is only advisable if some items are removed and the student scores are adjusted for DIF.  It was concluded that any DIF effects evident between the Danish and Swiss samples may be due to language differences, cultural differences, differences in the academic settings for the statistics courses, as well as combinations of these factors. We conclude that rigorous item analysis is a crucial pre-requisite for cross-cultural comparison studies utilizing psychological instruments to avoid misleading conclusions. Translation alone, no matter how rigorous, is not enough.