Henderik Alex Proper

Domain modeling is widely used in both academic and industrial contexts, yet the modeling process remains insufficiently understood, particularly with respect to the difficulties modelers experience in practice. Prior work has identified recurring modeling pain points and provided initial evidence of their prevalence, but offers limited support for systematically analyzing and addressing them in context. This paper extends previous research in two directions. First, it presents an empirical investigation based on a survey among 49 modelers, analyzing the frequency of modeling pain points, the frustrations they generate, the reactions they elicit, and the influence of external factors such as modelers’ experience, modeling language, and tool support. Second, it introduces an analytical method for characterizing modeling pain points at the level of individual instances. The method structures the analysis by explicitly relating context, underlying causes, associated frustrations, and the extent to which these causes can be addressed. The method was applied in two real-world cases, where it supported a more explicit understanding of how pain points arise and how they can be handled in practice. Taken together, the results contribute to a more systematic understanding of the modeling experience and provide directions for managing modeling challenges.

Context: Although research has advanced the understanding of modeling value, the resulting knowledge remains fragmented and difficult to apply in practice. Consequently, practitioners still lack structured approaches to assess and communicate the value of modeling, making decisions in modeling contexts difficult to justify and communicate, particularly in collaborative settings.Objectives: This study presents, demonstrates, and evaluates the Modeling Value Canvas, a theory-informed artifact designed to support the analysis and communication of the value and risk associated with a specific decision within a modeling context.Methods: The study follows a Design Science Research approach combined with Echelon Design Science Research. The artifact was developed through two design cycles. The first cycle involved four real-world case studies, while the second included an evaluation in an Ontology-Driven Conceptual Modeling course at the University of Twente, involving 10 completed Canvases and 29 qualitative user feedback responses. UTAUT2 guided the feedback collection, thematic analysis was used to analyze responses, and a scoring scheme was used to assess the completed Canvases.Results: The results indicate that the Canvas facilitated structured reflection, reasoning, and communication, even among novice modelers with no prior experience using the artifact. Participants were able to articulate value and risk considerations, stakeholder concerns, contextual conditions, and decision trade-offs. The redesign also reduced previously identified limitations related to conceptual clarity and usability.Conclusion: By making reasoning more explicit, traceable, and communicable, the Modeling Value Canvas supports more systematic analysis of decisions in modeling contexts. The artifact may contribute to strengthening modeling practice and fostering the adoption of modeling in organizations.