You are here: Home / Research / Research Lines / Methodological research and development

Methodological research and development

Research Line 5

Oliver Lüdtke (spokesperson), Gabriel Nagy, Aiso Heinze & Olaf Köller

As educational research and educational politics become more and more data-based, representative data collections in the school context have gained in importance in the last two decades. This can be seen, for example, in the Programme for International Student Assessment (PISA). As a result, the methodological and statistical challenges that educational research is confronted with are becoming increasingly complex. The datasets obtained by large-scale school assessment studies, for example, have a multilevel structure, which is due to the fact that students are nested within classes and classes are nested within schools. Statistical analyses are made even more difficult by the fact that data are often not available for all the individuals selected for a study (Missing Data) – as students either skip individual items in the study or do not participate at all.

Work in this research line concentrates on two areas: methodological research in education and psychology (educational measurement); and test development, test validation and educational monitoring (educational assessment).

This research area currently focuses on the following topics:

(1)  Educational measurement:

  • Further development of multilevel analytical procedures
  • Statistical modeling of competency structures and competency development
  • Missing data techniques
  • Estimating causal effects

(2)  Educational assessment:

  • Developing and validating basic education tests in mathematics, science and information technology
  • Experimental studies on the cognitive processes that take place during testing
  • Educational monitoring

These topics are closely connected to the working program of the Centre for International Student Assessment (ZIB) – both with regard to the research output (PISA reports) and the research topics in a narrower sense. Overall, the goal is to generate recommendations for research practice. Software (e.g., packages in R) will be developed to make procedures user-friendly.