The Problem With, And Answer To, MCQs
Throughout my career as a veterinary undergraduate, many of our exams have been in the Multiple Choice Question (MCQ) format, a method of assessment that, in my opinion, comes with considerable drawbacks. Generally, a question is asked, with 4 or 5 possible answers, of which one is correct. The first problem is one of testing depth of knowledge - with an MCQ, the correct answer is by definition already in front of the examinee, so simple questions of understanding are rarely asked, as they will be answered correctly by 99% of students. The questions asked are normally on very specific, and pretty detailed, pieces of knowledge. This contrasts to the challenges faced by a practicing vet - understanding a problem is absolutely vital, whereas exact facts and figures can always be looked up. As such, the MCQ is an inadequate method of assessing if the candidate has what is needed to achieve their ultimate aim, i.e. practicing as a veterinary surgeon.
The second problem is one of statistics. If candidates were to blindly guess at questions, they would, on average, answer one in four/five of these correctly, skewing the results of the exam. In order to combat this problem, many exams introduce “negative marking,” whereby 1/4 or 1/5 of a mark is deducted for every incorrect answer, with the stated aim of decreasing guessing. This is a totally ineffectual measure, as the student versed in statistics will realise that blindly guessing in the face of negative marking will yield neither a gain or loss of marks. So, if they have the slightest inkling of the answer, it is in their favour to go with it.
After having three (non-negatively-marked) MCQ exams last week, these thoughts have been buzzing round the old brain, and I’ve found a partial solution. Tell the students that the examination is negatively marked - this achieves the aim of discouraging outright guessing, which could unfairly skew some students results upwards. However, don’t actually apply negative marking - this allows students who were unfortunate enough to think they knew the answer, but were in fact wrong, to go unpunished, thus reducing the downward skew of results for unlucky students. Of course, this would rely on utmost secrecy about the non-application of negative marking - something probably quite difficult to achieve within an organisation as insular as a vet school!
As a grand finale, I present to you an idea that will most likely see me beaten to a pulp by my classmates! The best solution to the MCQ problem is, in my opinion, to use more traditional methods - short and long answer questions, with a choice of essays can test a wide range of subjects, to varying depths. Importantly, it also tests understanding, rather than just parrot fashion learning. It does of course require considerably greater effort to mark…which may or may not be a stumbling block to its reintroduction!


