Towards Improving Conceptual Modeling: An Examination of Common Errors and Their Underlying Reasons

Abstract

Databases are a critical part of Information Technology. Following a rigorous methodology in the database lifecycle ensures the development of an effective and efficient database. Conceptual data modeling is a critical stage in the database lifecycle. However, modeling is hard and error prone. An error could be caused by multiple reasons. Finding the reasons behind errors helps explain why the error was made and thus facilitates corrective action to prevent recurrence of that type of error in the future. We examine what errors are made during conceptual data modeling and why. In particular, this research looks at expertise-related reasons behind errors. We use a theoretical approach, grounded in work from educational psychology, followed up by a survey study to validate the model. Our research approach includes the following steps: (1) measure expertise level, (2) classify kinds of errors made, (3) evaluate significance of errors, (4) predict types of errors that will be made based on expertise level, and (5) evaluate significance of each expertise level. Hypotheses testing revealed what aspects of expertise influence different types of errors. Once we better understand why expertise related errors are made, future research can design tailored training to eliminate the errors.