As part of this master's thesis, an easily maintainable computerized adaptive testing platform was developed that can be extended to suit specific domains with little effort. Thanks to the optimized IT infrastructure, a large number of students can be assessed simultaneously without resource problems. The examples used for the so-called "SQL data modeling test" were exported from an existing online learning platform. Based on this data, 174 SQL test items were created and their difficulty levels were determined using a specially developed method. The matching of result tables obtained by database queries was realized by a line-by-line comparison method. Furthermore, a user test was used to identify and eliminate existing problems in the implementation and test process and to identify potential improvements for future implementation work.

Peter Baumann

Assoz. Univ.-Prof. Dipl. -Ing. Dr. Stefan Oppl