Statistical Evaluation of Multiple-Choice Assessments Development and validation of a statistical analysis method to monitor the European Ophthalmology exit examination, and evaluation of examination questions and methods

The validity of Type X MCQ's for assessment purposes has been doubted by many lucationalists because of poor statistical (item) performance parameters. However, these (item) performance parameters can be enhanced by implementing negative marking for incorrect answers. The same educationalists claim that negative marking for incorrect answers is not advisable since it would discriminate against female candidates. However, it seemed impossible to find any trace of published objective data supporting this widely cited hypothesis.;One of the main arguments to use type X MCQ's for the European Ophthalmology exit examination, is the multilingual character of the assessment: Type A MCQ's) which are advised by the earlier mentioned educationalists, may indeed be vulnerable to answer shifts due to translation issues.;Another argument supporting this choice is the fact that by using Type X MCQ's it is also possible to assess candidates who have partial knowledge on the subject.;Finally, doubt can be casted on the argument that type X MCQ's would be more difficult to construct compared to type A MCQ's, as from personal experience has learned that for both MCQ-types, it is essential to invest time and energy on writing good MCQ's.;Being convinced that type X MCQ's have validity for assessment purposes, the major aim of this doctoral research project concerned the evaluation of (item) performance parameters and the investigation whether indeed negative marking for incorrect items would be discriminating towards female candidates.