Multiple choice questions = box-ticking questions
Today, the testing of medical students' knowledge often takes the form of a digital exam, i.e. with the aid of a computer. Based on experience from the United States and other countries, so-called multiple choice questions, often abbreviated MCQ, are increasingly used (1, 2). Such questions are provided with multiple response alternatives, usually from three to five, of which only one is the 'single best answer'. The other response alternatives are referred to as distractors. These alternatives need to be seen as plausible, should not stand out or be glaringly wrong, but still less correct than the 'single best answer'. Multiple response questions are a variant of multiple choice questions, where the candidates must select two to three correct responses from among five to eight alternatives. Other, less frequently used question types also exist.
The main advantage of such box-ticking questions is that they can be answered within a short time. An exam that involves only such questions can therefore include more questions than other types, and thereby encompass larger elements of the discipline. In addition, the scoring is done automatically, since only one response alternative is the 'single best answer' (or more in multiple response questions). By starting with a description of a clinical situation, the exam question can simulate a clinical decision-making process (4). Sets of good multiple choice questions have been shown to differentiate well between strong, intermediate and weak students (2).
Students learn to identify the correct answer by looking at the wording of the questions – they become test wise
However, good multiple choice questions are difficult to formulate, and not all topics are equally well suited for them. The criticism raised against these types of questions also focuses on their failure to reflect clinical realities, that they do not adequately test the candidates' ability to reflect and apply their knowledge, and that they have a negative effect on the students' learning behaviour (2, 5). Moreover, the likelihood of simply guessing the correct answer is high: 25 % for four response alternatives and 33 % and 50 % respectively if the candidate is able to identify one or two incorrect response alternatives. Some candidates will recognise the correct answer when reading the response alternatives given, referred to as cueing. Publishing of previous exam questions and experience from sitting previous exams will help students to learn to identify the correct answer by looking at the wording of the questions – they become test wise. These disadvantages of multiple choice questions are often underestimated and undercommunicated.