An evaluation of novice end‐user computing performance: Data modeling, query writing, and comprehension

Abstract

End-user computing has become a well-established aspect of enterprise database systems today. End-user computing performance depends on the user-database interface, in which the data model and query language are major components. We examined three prominent data models-the relational model, the Extended-EntityRelationship (EER) model, and the Object-Oriented (OO) model-and their query languages in a rigorous and systematic experiment to evaluate their effects on novice end-user computing performance in the context of database design and data manipulation. In addition, relationships among the performances for different tasks (modeling, query writing, query comprehension) were postulated with the use of a cognitive model for the query process, and are tested in the experiment. Structural Equation Modeling (SEM) techniques were used to examine the multiple causal relationships simultaneously. The findings indicate that the EER and OO models overwhelmingly outperformed the relational model in terms of accuracy for both database design and data manipulation. The associations between tasks suggest that data modeling techniques would enhance query writing correctness, and query writing ability would contribute to query comprehension. This study provides a better and thorough understanding of the inter-relationships among these data modeling and task factors. Our findings have significant implications for novice end-user training and development.

The effects of database complexity on SQL query formulation

“…Additionally, the effects of task complexity (Topi et al, 2005) and data model representation (Chan et al, 1997(Chan et al, , 2005Rho and March, 1997) on query formulation have been studied. However, given that there exists no scientific evidence regarding the effects of logical complexity of a relational database on SQL query learning, we address here studies that consider database complexity in education in general.…”

Evaluation of information retrieval systems using structural equation modeling

“…In addition to the investigation of socio-economic phenomena, some uses of SEM regarded research areas that are somehow relevant to IR, since the factors affecting users' access to information systems were investigated. SEM was utilised by Chan et al [2005] to examine the multiple causal relationships among the performances for different tasks (modeling, query writing, query comprehension) performed by the users of a database interface, in which the data model and query language are major components. A structural equation model was also used to investigate users' behaviour within community networks 4 by Kwon and Onwuegbuzie [2005], Bulletin Board Systems by Chen and Chiu [2007], Wikipedia by Cho et al [2010], social network systems by Kipp and Joo [2010] and Park [2014], electronic commerce by Lu and Zhu [2010] and Afzal [2013], library systems by Sin [2010], exploratory search by O'Brien and Toms [2013], agile software development by Senapathi and Srinivasan [2014], and online education by Zhang and Dang [2015].…”

An evaluation criterion for open source software projects: enhancement process effectiveness

“…To do that, a step-by-step process suggested by Hair et al (2006) was applied. PLS-Graph version 3.00 (Chin 2001) was used to conduct CFA (Sun & Zhang 2008;Chan et al 2005;Lee et al 2007). Bootstrap resampling procedure with the number of samples 200 was also employed to test the significance of all paths (Cotterman & Senn 1992).…”