Background: Progress testing is a form of longitudinal examination which, in principle, samples at regular intervals from the complete domain of knowledge considered a requirement for medical students on completion of the undergraduate programme. Over the course of the programme students improve their scores on the test, enabling them, as well as staff, to monitor their progress.
Aim: We aimed to review methods which have been used to assess the results of individual tests, and to make recommendations on best practice.
Discussion: In assessing progress tests, there are a variety of choices that must be made. These include whether the test is norm- or criterion-referenced; whether marking is negative or "number-right"; whether the grades are reported on a continuous or a discontinuous scale, and whether the grades are weighted towards the most recent observations, or the entire set of grades is used to determine the final grade. Grade boundary setting in the context of progress tests is also considered, using a mathematical model to predict the consequences of different approaches. The relationships between boundary setting, progression and remediation rules are considered.
Conclusions: We concluded that norm referencing is preferable to criterion referencing, negative marking preferable to number-right marking, a discontinuous scale preferable to a continuous scale and that grades should be weighted to favour the most recent outcomes, although there should still be a degree of persistence (earlier grades should not disappear all together). Grade boundaries should be established with regard to rules on remediation and progression.