Abstract: The thesis examines how project data from different Qualitative Data Analysis (QDA) tools can be exchanged through a common domain model. It asks which entities and relations recur across the analysed formats, which of them should be represented in the model, and where conversions remain lossy. The method combines project-file and documentation analysis with reverse engineering, format-specific mapping, and automated round-trip tests. The result is the Corpus model and a TypeScript library, which provides paths for reading and writing REFI-QDA (.qde and .qdpx), MaxQDA (.mx20, .mx22, and .mx24), ATLAS.ti (.atlasproj), and NVivo (.nvp) artifacts (QDArchive, 2026). The evaluation indicates that a recurring core of entities and relations can be represented in a common model for the tested properties, while target formats differ in their semantic coverage. Structured diagnostics and optional strict export failure make these differences visible.
Keywords: QDA, TypeScript, domain modeling, reverse engineering, qualitative data analysis
PDF: Bachelor Thesis
Reference: Benedikt Zinn. A Common Domain Model for QDA Artifacts. Bachelor Thesis. Friedrich-Alexander-Universität Erlangen-Nürnberg: 2026.
Discover more from Professorship for Open-Source Software
Subscribe to get the latest posts sent to your email.