On the Syntactic, Semantic, and Pragmatic Quality of Students’ Conceptual Models

AbstractVisual notations and conceptual models, such as ER diagrams or UML diagrams aid in aligning stakeholder needs, defining and prioritizing processes and goals for the system under development, serve as a reference for requirements elicitation, negotiation, and enable validation as well as verification of artifacts. With such a ubiquitous presence and paramount importance, conceptual models have therefore been introduced in software engineering curricula far and wide. However, it is exceedingly difficult to teach and learn conceptual modeling. Not only does it require educators to instruct notation and syntax of the visual language, but also semantic intricacies. Similarly, students struggle with what differentiates a “good” conceptual model from an inadequate one, how to use conceptual models of different types in conjunction with one another in a meaningful way, or simply how to avoid ambiguity and vagueness. In this paper, we discuss the syntactic, semantic, and pragmatic quality of conceptual models in four courses from an undergraduate software engineering program. It is not our aim to present empirically rigorous results, but to contribute to the body of knowledge on the quality of typical novices’ conceptual models. We seek to foster discussion in the community and present observations and results for comparison.

Return to previous page