Abstract
Multiple-choice questions (MCQs) are a common form of assessment in medical science education. The traditional MCQ format involves students picking a single best answer (SBA) from four or five options. There are concerns about the ability of SBA formats to reward partial knowledge and their susceptibility to guessing. An alternative to SBA is elimination testing (ET), wherein students eliminate all the incorrect answer options, with negative marking to deter guessing. Cognitive load theory (CLT) is an approach to education that prioritises strategies to minimise the amount of unnecessary 'load' placed upon working memory. The cognitive load imposed by assessment design has received little attention. We evaluated the cognitive load of SBA and ET MCQ formats, using an online participant pool and a survey of students at a UK Medical School. We found that partial knowledge was rewarded with the ET format. However, students strongly preferred the SBA format and reported both that it was easier and imposed a lower cognitive load. Removing negative marking reduced the cognitive load of ET style questions and improved the student experience, but the improvement was insufficient to change student preference for SBA questions.