This empirical research study published in the journal Frontiers in Psychology in October 2018 examined why approximately three-quarters of people (on average) reached incorrect answers when presented with statistical reasoning problems using natural frequencies, despite this format being ostensibly more intuitive than probability format. Through their experimental study focusing on cognitive processes, the researchers found that the majority of their 180 participants chose to translate natural frequencies into probabilities before attempting to solve the problem, therefore "reveal[ing] the Einstellung effect in Bayesian reasoning situations," in which "people often rigidly apply a previously learnt solution strategy while neglecting possibly important information that would allow an easier solution." The authors note that since "[m]any professionals, such as medical doctors and judges in court, are expected to make momentous decisions based on statistical information," the real-world consequences of the Einstellung effect in this type of scenario can be severe, such as misdiagnoses or false convictions. The lead author of this study is Patrick Weber, a doctoral student in mathematics education at the University of Regensburg (Germany), and the full study, written in accessible language, is publicly available.
Comments