ABSTRACT:

Different types of conceptual database modeling representations have been developed to construct conceptual database models. Each data modeling representation uses a different form for representing entities, data items, and associations. This article reports the results of an experiment that examined two conceptual database modeling representations, the entity-relationship representation and the relational representation, to determine their effectiveness in modeling and documenting accounting databases. The experimental task required subjects to carry out an audit review of conceptual data models of five common accounting transaction processing cycles.

The results indicate the superiority of the entity-relationship data modeling representation over the relational representation in facilitating the audit review of accounting database conceptual models. This superiority, which is due to the graphical nature of the entity-relationship representation, suggests that using the entity-relationship conceptual modeling representation for database documentation will result in more effective audits in accounting database processing environments.

The conceptual database model is a representation of the organization's database that helps designers, users, and auditors understand the contents of the database. The conceptual model plays a significant role in the auditor's review of processing (Vasarhelyi and Lin 1988; IIARF 1991). The IIA notes that the auditor should review conceptual database models to verify that they represent the data accurately and completely (IIARF 1991, 5-48). If the conceptual database model erroneously represents the organization's transaction processing, that would indicate to the auditor that the underlying database implementation is inaccurate. Conceptual modeling errors would, in turn, signify that there are errors in database processing.