Applications of multi-level models to research policy and higher education studies

This multilevel short course is aimed at providing an introduction to multi-level modeling as a research method for RISIS researchers and give them the tools to apply the method to their work. The course in divided into a statistical teaching of the method as well as an guided application through a practical group exercise. The former part focuses on the rationale for multi-level modeling and how it can be applied to research in the field the higher education. The latter part allows for a hands-on guidance with using the technique with RISIS datasets and allows for participants apply it to different settings and discuss the interpretation of the results and its limitations.
Multilevel modeling gives researchers substantial leverage in overcoming the usual problems associated with either ignoring multiple levels of data or combining them.
The inherent structure of higher education and research policy data makes it well suited for multi-level modeling. For instance, researchers tend to be nested into laboratories within universities or in comparative analysis universities are nested into countries. In order to demonstrate the analytical value for the research policy field, the course will include demonstrations and practical exercises done with datasets available in RISIS.