Improving User Performance in Conditional Probability Problems with Computer-Generated Diagrams

摘要

Many disciplines in everyday life depend on improved performance in probability problems. Most adults struggle with conditional probability problems and prior studies have shown user accuracy is less than 50%. This study examined user performance when aided with computer-generated Venn and Eu-ler-type diagrams in a non-learning context. Following relational complexity, working memory and mental model theories, this study manipulated problem complexity in diagrams and text-only displays. Partially consistent with the study hypotheses, complex visuals outperformed complex text-only displays and simple text-only displays outperformed complex text only displays. However, a significant interaction between users' spatial ability and the use of diagram displays led to a reversal of performance for low-spatial users in one of the diagram displays. Participants with less spatial ability were significantly impaired in their ability to solve problems with less relational complexity when aided by a diagram.