Possibly the only arguments in favour of assessing statistical understanding by traditional written examinations may be that this is one of the very few forms of assessment where we can be sure that the work is the candidate’s own, and that, as they are used extensively in higher education and by professional bodies, similar examinations at school level can be seen to be preparation for post-school examinations. Arguments that are often made against this form of assessment, however, stress its artificial nature and the unfairness that the time constraint may cause for some candidates. Examinations usually test only the relatively short-term memory and there is no guarantee that facts learned for the test will be retained.