Abstract
This study analyzed university students' reasoning modes in the vaccination problem (Itoh, 2006, 2007) in partitive formulation (Macchi, 1995, 2000) and the cab problem (Tversky & Kahneman, 1980) in nonpartitive formulation. Both problems comprised a basic first-order quantification of probability problem, a second-order quantification of probability problem with additive composition, and a Bayesian problem. The proportion of correct answers for the vaccination problem was significantly larger than that for the cab problem. This reflects the effect of partitive formulation. The reason an analysis of the nature conducted in the basic first-order quantification of probability problems and the second-order quantification of probability problems with additive composition in this study has not been carried out previously is that studies have only focused on the Bayesian reasoning. Further studies using content-free problems can illuminate the fundamental properties of a cognitive system in probabilistic reasoning.
