Hattori and Nakagawa (1996) and Hattori (2000) proposed an experimental method to identify the logical reasoning processes of human beings. This study showed that their method is not only useful for identifying the processes, but also valid for a learning support system for logical reasoning, especially for conditional inference. The experiment examined the validity of the method. The results showed a significant increase in the proportion of correct answers in the experimental group only, and the effectiveness of the method was demonstrated. Furthermore, with reference to the subjects' verbal data, a variety of features in the learning processes were identified and examined. The discussion includes an examination of the indirect learning processes, the drastic change in performances, the time-lag before the emergence of the effects on learning, the effects of confusion and mislearning, and the maintenance of apparent consistency using individual local rules.
View full abstract