The working unit Educational Measurement deals with the measurement and analysis of psychological characteristics which are of particular importance for applied education research. This includes, most notably, questions regarding the measurement of latent (unobservable) characteristics, and research on psychological methods in large scale assessments.
Measuring latent characteristics:
Psychological constructs are, as a rule, not directly observable. Instead, conclusions regarding the actual latent characteristics of interest must be made based on observed answers (to test questions, for example). In order for tests which measure psychological characteristics to be useful, they have to meet certain standards of test psychology. On the basis of psychometric test models, this working unit examines the quality of the test items (the competency tests in particular) used by NEPS, and creates personal scores as as estimate of the specific latent characteristics of interest.
Research on methods of large scale assessment:
The longitudinal observation of psychological characteristics in different age cohorts represents a particular challenge for survey design and the scaling of the constructs thus surveyed. Based on analyses of the tests used in NEPS, this working unit develops longitudinal models of competency development and questions for linking and dealing with missing values. There is also a focus on the development of test designs for large scale assessment, for instance the form of adaptive computer surveys.
These developments and other current research take place in close collaboration with the working units for Competencies and Statistical Survey Methods.
Further assigned to the working unit Educational Measurement are the following projects:
Scaling and Test Design