![]() |
|
||||||||||||||||||||||
|
Yuri Sato The University of Tokyo Yuichiro Wajima The University of Tokyo Kazuhiro Ueda The University of Tokyo
We analyze the information discrepancy between diagrammatic representations and logical reasoning, which we call visual biases in diagrammatic reasoning. Diagrammatic representations contain semantic information, which is based on the topological configurations of objects, and visual information, such as geometric location. In principle, visual information is unnecessary to the validity of logical reasoning. However, people are so sensitive to visual information such as size and shape in diagrams that they occasionally do not ignore irrelevant information. This phenomenon leads to mistakes in logical reasoning. We addressed this issue in the present study. In Experiment 1, we assessed whether and how a visual bias of external diagrams affects reasoning performance. We asked participants to directly manipulate size-fixed (Euler) diagrams while solving syllogistic tasks. In Experiment 2, we tested whether size-scalable diagrams were able to reduce a visual bias of diagrams in logical reasoning.
Visual bias of diagram in logical reasoning (59 KB)