UXARclass: A Proposed Classification Taxonomy of Design Recommendations Toward Relevant UX Dimensions for Augmented Reality in Corporate Training
Stefan Graser, Martin Schrepp, Stephan Böhm · International Journal of Human-Computer Interaction · 2026
Augmented reality (AR) in Corporate Training (CT) enables innovative, immersive training scenarios, creating a new User Experience (UX). AR authors, therefore, apply design recommendations for application development to convey a specific UX. However, it remains unclear how the specific AR design recommendations contribute to the respective UX, typically measured by applying standardized questionnaires. Users may perceive it differently, leading to a misalignment between the intended and actual UX. In turn, quantitative evaluation results regarding UX questionnaires are often too general and not useful in practice, as they capture a high-level impression. This highlights the gap between design practice and users’ subjective impressions. Therefore, linking design recommendations with the relevant UX dimensions enables an innovative, evidence-based approach, allowing a more precise interpretation of results, providing useful suggestions for system improvement. By systematically aligning design recommendations to the relevant UX dimensions, we aim to bridge the gap between quantitative measurement and qualitative design actions. We present a Delphi-based multi-method approach comprising three stages, each with a separate study. As a result, we provide a classification taxonomy, UXARclass, providing an assignment of AR design recommendations to the relevant UX dimensions regarding the AR-specific UX questionnaire, UXARcis.