The probabilistic reasoning in "real-world" context: some reflections about the judgement in medical problem solving

This work aims to assess the probabilistic reasoning applied in the "real-world" context, specifically referring to the judgement in medical problem solving, applied by unexperienced individuals. It was carried out a mixed methods approach, in which data collected by quantitative methods was integrated with data collected in qualitative method. The open response to the probabilistic medical problem was assessed and classified using the structure of the observed learning outcome (SOLO) taxonomy. This assessment was associated with quantitative data related to the correctness of the response, with the confidence in the response, and with numerical and visuospatial abilities assessed using Thurstone's Primary Mental Abilities (PMA) scales. The data were collected while administering the problem to 352 Italian undergraduates in psychology, who had no statistical knowledge. Few individuals gave the correct answer and few applied high levels of reasoning. Furthermore, many exhibited low levels of confidence in the correctness of their responses. It was found that the participants often applied avoidance strategies. The implications of these findings with respect to applied perspectives will be discussed.