Since 2008, electronic examinations have been conducted at the University of Veterinary Medicine Hannover, Germany which are analyzed extensively in the current study. The aim is to assess the quality of examinations, the status quo of the electronic examination system and the implementation of recommendations regarding the conduct of exams at the TiHo. Based on the results suitable indicators for the evaluation of examinations and items as well as adequate quality assurance measures and item formats are to be identified. For this purpose, 294 electronic examinations carried out from 2008 to 2022 of the veterinary medicine course with an average of 248 participants each were evaluated with regard to the quality criteria reliability, difficulty index, and discrimination index. The main finding was that the number of items and the proportion of reused questions were identified as factors through which the quality of the examinations can be increased with simple adjustments. A higher number of items led to better reliability, whereby the required minimum reliability in examinations of 0.8 was reliably achieved from an item number of 98 questions. The proportion of reused questions should be kept low, as these had a negative influence on the characteristic values. Measures accompanying examinations, such as training of question authors and a pre- and post-review process, should also ensure the quality of examinations. For the post-review process, the distribution of examination results, reliability, item and distractor analysis are adequate indicators for evaluating examinations.
Keywords: Cronbach's a; discrimination index; e-assessment; examinations; item formats; multiple choice questions (MCQ); veterinary education.