Marks across lower secondary schools in Norway: What can be explained by the composition of pupils and school resources
The aim of this report is provide background information for how to construct informative performance indicators for schools at the Norwegian lower secondary education level (”Ungdomsskolen”) based on pupils' achievement as measured by their marks. It is commonly accepted that "school quality", however defined, may have great importance for how much pupils learn. The question is whether it is possible to quantify how schools differ in their contribution to pupils' learning. There are many other factors than the school itself, such as pupil composition and resource use, that can possibly explain differences in marks across schools. Whether and how to adjust for such factors when comparing school performance is not obvious. The relevant adjustment procedure depends on the question(s) asked and availability of data. Construction of such measures places great demands on the data, and the ideal solution may not always be feasible. This report analyses these issues in more detail, with particular focus on how reliable school performance measures may be constructed given the present availability of data in Norway. We focus our discussion around the following questions: Do differences in marks between schools reflect "real" differences or random noise? What is the impact of family background on the school results of individual pupils and differences between schools? Are differences between schools statistically significant? Do resources at schools have an impact on the performance of pupils?
Hægeland, T., Kirkebøen, L.J., Raaum, O. og K.G. Salvanes