Title: Score-Tests for Detecting Differential Item Functioning in Cognitive Diagnosis Models Authors: Michel Philipp (1), Achim Zeileis (2), Carolin Strobl (1) Affiliation: (1) University of Zurich, (2) University of Innsbruck Abstract: Cognitive diagnosis models (CDM) are a family of psychometric models for analyzing dichotomous response data. They provide detailed information about mastery or non-mastery of predefined skills, which are required to solve the tested items, and can thus reflect the strengths and weaknesses of the examinees in the form of a skills profile. The simplest version of a CDM is the non-compensatory DINA model with only two parameters per item. Despite their popularity in psychometrics, differential item functioning (DIF) is relatively unexplored for CDMs and most models are not designed to incorporate covariate effects. In this talk, we will present a brief literature overview as well as the current state of our research on a score-test for detecting DIF in the DINA model. Score-tests allow for testing parameter differences between given groups suspected of DIF, such as males and females, but are also computationally feasible for detecting parameter differences along continuous variables, such as age, without previous discretization. Further steps of this work in progress will be to compare the new score-test to standard DIF detection approaches for CDMs in simulation studies and explore its integration in the model- based recursive partitioning framework.