Readers may have noticed new text on the inside front cover of Advances. Psychiatrists participating in the American Medical Association (AMA) CME system can now claim credits by completing the MCQs that feature at the end of each article. Although MCQs are not universally loved, the attractions of using them for summative assessment are significant. Well-written MCQs are reliable, test breadth of knowledge and are easy to administer and mark. Certificates generated through the College's CPD Online website (www.psychiatrycpd.co.uk) give evidence of participation in CPD and CME programmes. Such evidence, held in portfolios, is kept for regulatory processes such as appraisal and revalidation.

MCQs can also be used for personal development. In the first issue of Advances the Editor wrote: ‘There will be opportunities for the reader to test what has been learnt from each article in the form of a few multiple choice questions’ (Reference SimsSims 1994). This in a nutshell is formative assessment – a personal test of understanding rather than a formal evaluation of knowledge. Answering MCQs gives feedback in self-assessment, checking for yourself rather than proving to someone else. Not knowing the best answer reveals a gap in knowledge. Getting a question wrong may be more useful still. ‘It is impossible for a man to learn what he thinks he already knows’, observed the Greek Stoic philosopher Epictetus (c. 55–135 ad). The disquiet experienced in finding out you are wrong when sure you are right, cognitive dissonance, is at the root of all learning. It has been called ‘learning-by-disturbing’ (Reference Robinson, Purves and WilsonRobinson 2003). And testing supports not only learning but also remembering. We sense that repeatedly reading an article is a good strategy to aid recall. But we are deceived. Testing has a powerful positive effect on long-term retention which re-reading does not (Reference Roediger and KarpickeRoediger 2006). What follows is that, to be helpful for learning, MCQs ought to focus on the key points of an article. Trivial, ambiguous or superficial questions (and answers) are unhelpful. Authors should use questions to signal their take-home messages.