Analysis of Competing Data Structures: Does Ontological Clarity Produce Better End-User Query Performance?

End users respond to stakeholders' information requests by using query tools to retrieve information from their organizations’ data stores. The structure of these data stores impacts end users’ performance, e.g., the accuracy of their responses. Ontologically clearer conceptual models have been shown to facilitate better problem solving within real-world application domains. If, however, ontologically clearer conceptual models are directly transformed into implementation (logical) data models, the differences in the number of entities and relationships may cause cognitive issues for end users that are likely to affect their query performance. This paper reports the results of an experiment that investigated the effect on query performance of more traditional logical models compared to ontologically clearer logical models. Results indicate that end users of the 1 Jeffery Parsons was the accepting senior editor. This paper was submitted on 28 February, 2005, and went through 2 revisions. Analysis of Competing Data Structures/Bowen, O’Farrell, & Rohde Journal of the Association for Information Systems Vol. 7 No. 8, pp. 514-544/August 2006 515 ontologically clearer implementation made fewer semantic errors overall. Thus, the benefits of ontological clarity at the conceptual level may translate into similar benefits when querying ontologically clearer logical models. Unfortunately, an examination of the specific types of errors that were made indicated that the benefits are not clear cut. While the removal of optional attributes and relationships led to an overall reduction in the number of errors, closer analyses show that some types of errors (involving projection and restriction) decreased as expected, while other types of errors (involving joins) increased.