ZIB Education

Item Fit – Statistical and Practical Significance of Item Misfit in Educational Testing

Lead: DIPF
Joint Research with: TUM, IPN, Boston College

Project management: Dr. Carmen Köhler➚Prof. Dr. Johannes Hartig➚

Scientific staff: Katharina Fährmann➚

Project cooperations: IPN (Dr. Alexander Robitzsch➚), Lynch School of Education and Human Development (Prof. Dr. Matthias von Davier➚), TUM (Dr. Jörg-Henrik Heine)

The project Item Fit – Statistical and Practical Significance of Item Misfit in Educational Testing, funded by the Deutschen Forschungsgemeinschaft (DFG), isa data collaboration between ZIB and DIPF on PISA field test data. The project aims to derive guidelines for the use of item fit statistics in large-scale educational studies. The focus is on the practical significance of items whose response differs significantly from the response probabilities predicted by the model. The effects on the estimation of competences by or without the inclusion of statistically unsuitable items will be investigated.

The core objective of the data cooperation is to validate methods for assessing the practical significance of items that do not fit the model using empirical field test data from the PISA 2018 survey.

The contact person for data collaboration at ZIB is Dr. Jörg-Henrik Heine (location TU Munich).

Further information on the project can be found here



Investigation of responses that differ significantly from the predictions. © pexels.com


089 289 28274 zib.edu@sot.tum.de