Determining influence, interaction and causality of Contrast and Sequence effects in OSCEs.

In this instance, the small DRIFT effect we observed on embedded video can be causally attributed to examiner behaviour. Contrast effects appear less ubiquitous than some prior research suggests. Possible mediators of these finding include: OSCE context, detail of task specification, examiners' cognitive load and the distribution of learners' ability. As the operation of these effects appears to vary across contexts, further research is needed to determine the prevalence and mechanisms of contrast and DRIFT effects, so that assessments may be designed in ways which are likely to avoid their occurrence. Quality assurance should monitor for these contextually variable effects in order to ensure OSCE equivalence.