Empirical validity for a comprehensive model on educational effectiveness

Educational effectiveness research is often criticised because of the absence of a theoretical background. In our study we started out from an educational effectiveness model which was developed on the basis of educational theories and empirical evidence. We have tested the main assumptions of the model by reanalysing a large-scale longitudinal dataset which contains elementary school data of subsequent student cohorts, their teachers and their schools. We found some expected positive effects on achievement of individual classroom and school factors in the effectiveness model but these were not stable across school subjects and student cohorts. When factors were grouped into clusters the findings remained essentially the same. To interpret the clusters we made use of the concept of consistency, which refers to a balance between factors which may enhance effectiveness. Some teachers and schools are more consistent than others, but this does not affect the achievement of their students. Cross-level interaction effects were not found. The instability of effects makes it hard to suggest changes in the model at this moment.