ContextSome characteristics of assessments exert a strong influence on how students study. Understanding these pre-assessment learning effects is of key importance to the designing of medical assessments that foster students' reasoning abilities. Perceptions of the task demands of an assessment significantly influence students' cognitive processes. However, why and how certain tasks positively drive' learning remain unknown. Medical tasks can be assessed as coherent meaningful whole tasks (e.g. examining a patient based on his complaint to find the diagnosis) or can be divided into simpler part tasks (e.g. demonstrating the physical examination of a pre-specified disease). Comparing the benefits of whole-task and part-task assessments in a randomised controlled experiment could guide the design of assessments for learning'. ObjectivesThe purpose of this study was to determine whether the knowledge that an objective structured clinical examination (OSCE) will contain whole tasks, as opposed to part tasks, increases the use of diagnostic reasoning by medical students when they study for this assessment. MethodsIn this randomised, controlled, mixed-methods experiment, 40 medical students were randomly paired and filmed while studying together for two imminent physical examination OSCE stations. Each 25-minute study period began with video cues and ended with a questionnaire on cognitive loads. Cues disclosed either a part-task OSCE station (examination of a healthy patient) or a whole-task OSCE station (hypothesis-driven physical examination [HDPE]). In a crossover design, sequences were randomised for both task and content (shoulder or spine). Two blinded and independent authors scored all 40 videos in distinct randomised orders, listening to participants studying freely. Mentioning a diagnosis in association with a sign was scored as a backward association, and the opposite was scored as a forward association; both revealed the use of diagnostic reasoning. Qualitative data were obtained through group interviews. ResultsStudying for whole-task OSCE stations resulted in a greater use of diagnostic reasoning. Qualitative data triangulate these findings and show the precedence of cues sourced from the student grapevine'. ConclusionsIn comparison with traditional' part-task OSCEs, whole-task OSCEs like the HDPE increase students' use of diagnostic reasoning during study time. Discuss ideas arising from the article at discuss.