The publication of PISA has affected both education research and education debate immensely. Thereby the education differences within Germany are assessed very differently: Often singular factors as migration background, social inequality or education policy are mentioned as reasons and the PISA-studies are politically instrumentalized. Given the correlation of migration background and social inequality, especially in a country of immigration like Germany it raises the question of the causal mechanisms. This article examines the relative explanatory power of these factors using empirical analyses.

On the basis of PISA 2006 we analyse determinants of science competence in Germany by using regression analysis. As a result, it can be shown that migration background and social classes not only have an impact on the competence each, but the interaction of both has an additional effect. Furthermore, there are significant regional disparities which are rarely examined due to data-specific problems. The juxtaposition of OECD-country comparison with comparison on German state level demonstrates the danger of drawing false conclusions because of the whole-nation bias. The analysis emphasizes the strong impact of the research design for empirical results and its resulting political implications.