The impact of alternative diagrams on the accuracy of recall: A comparison of star-schema diagrams and entity-relationship diagrams

摘要

Data warehouses can be constructed using either a relational or a dimensional model. This research compares the diagrammatic representations of these two models. Semantic network theory suggests that, because of their structure, star-schema diagrams (SSDs) are more understandable than entity-relationship diagrams (ERDs). Through two experiments that test both the accuracy and pattern of users' recall, we find strong evidence that SSDs are easier to understand than ERDs. Our findings imply that the use of the dimensional model should help decision makers query the warehouse directly, thereby reducing costs and the need for intervention and support by IT professionals.