Can positive expectations help to improve the learning of risk literacy? A cluster-randomized study in undergraduate medical students

BACKGROUND: Risk literacy, i.e., the ability to calculate and apply risk parameters, represents a key competence for risk communication and medical decision making. However, risk literacy is reportedly low in medical students. The successful acquisition of statistical competencies is often difficult, and can be hampered by emotional learning obstacles, calling for interventions to support learning. In this cluster-randomized study, we aimed to translate findings from placebo research to medical education. Specifically, we tested if the acquisition of risk literacy during a seminar unit can be facilitated by positive expectations, induced by a positive and non-threatening framing of the content and learning goals. METHODS: The study took place during a mandatory 2.5-h seminar on “risk literacy” for 2(nd) year medical students. The seminar teaches both statistical knowledge and its application in patient communication. To test the effects of expectations on risk literacy acquisition, the (otherwise identical) seminar was framed either as “communication training” (positive framing condition) or “statistics seminar” (negative framing condition). All N = 200 students of the semester were invited to participate, and cluster-randomized to the positive or negative framing condition (4 seminar groups each condition). Risk literacy was assessed with the “Quick Risk Test” (QRT) at the beginning and end of the seminar, along with statistics anxiety and subjective learning success using questionnaires. RESULTS: Data from N = 192 students were included. At the end of the seminar, risk literacy was increased in both framing conditions, with a significantly greater increase in QRT scores in the positive framing condition. Statistics anxiety was significantly decreased in both framing conditions, with no evidence of group differences. Subjective learning success was overall high and comparable between groups. CONCLUSIONS: Supporting our hypothesis, positive framing led to a significantly greater increase in risk literacy (i.e., in QRT scores). Our data offer first support that positive framing of learning goals may help to facilitate the acquisition of statistical knowledge. Expectation-orientated interventions may thus offer a feasible tool to optimize learning settings and framing of learning objectives in medical statistics courses.