Buonocore, Sara (2026) INVESTIGATING HUMANS WHILE LEARNING THROUGH EXTENDED REALITY TECHNOLOGIES: TOWARDS ADAPTIVE TRAINING. [Tesi di dottorato]

Documento PDF SARA__BUONOCORE_PHD_THESIS.pdf Visibile a [TBR] Amministratori dell'archivio Download (35MB) | Richiedi una copia

Abstract

Extensive research has shown that Extended Reality (XR) technologies can effectively support training in complex industrial contexts where real-world practice is risky, costly, or operationally disruptive. As the demand for more efficient and personalized learning solutions grows, adaptivity is recognized as a key requirement to ensure that guidance, pacing, and interaction align dynamically with trainees’ evolving performance and needs. However, when translating strategic interest into concrete interventions, organizations and instructional designers still face significant challenges, lacking clear, evidence-based guidelines for designing and validating adaptive XR training systems. This thesis addresses this gap by analyzing: 1) unexplored relationships between specific design features of XR environments and XR-related Human Factors; and 2) the underexplored modeling of users’ profiles to better understand how user characteristics moderate performance and experience in XR environments. The novelty of this work lies in the reconceptualization of adaptivity. Here, for the first time, it is modeled as a two-level property, including both profile-based pre-adaptation and performance-based dynamic adaptation. Further, the proposed methodology incorporates also XR training systems’ non-adaptivity, that is, the system’s set of core features, in one integrated approach. Alongside its methodological contributions, the thesis serves as a practical decision-support tool to guide industrial stakeholders in building their chain of evidence, and design and validate multi-level adaptive XR training features, based on measurable impact. Its applicability is demonstrated in two industrial case studies and further extended to a scientific outreach scenario in nuclear fusion, supporting transferability across learning domains.

Downloads

Actions (login required)

Modifica documento