Data Mining and Analytics in the Context of Virtual Reality: Benefits and Affordances

• In Event: Making Constructivist Learning Feasible: Exploring Convergence of Virtual Reality and Learning Analytics for Future Research Directions

Abstract

Current analysis of data from virtual reality (VR) environments is mostly done in retrospective, i.e., big data available from VR settings are being analyzed for identifying phenomena and/or predicting possible future states of learning processes or learning outcomes. In contrast, learning analytics focus (near) real-time analytics and direct support of educational phenomena (Ifenthaler, 2022), for instance adapting learning artifacts for supporting learning processes or providing personalized recommendations for achieving a specific competence level. This contribution will focus on the benefits and challenges of learning analytics with a specific focus on virtual reality environments.

Learning analytics are defined as the use of static and dynamic data about learners, and educators, as well as their context and learning environments, for real-time modeling, prediction, and support of learning processes, learning environments, as well as educational decision-making (Ifenthaler, 2015). From a holistic perspective, educational organizations and involved stakeholders can derive multiple benefits from learning analytics by using different data analytics strategies to produce summative, real-time, and predictive insights and recommendations which can be associated with four levels of stakeholders: mega-level (governance), macro-level (organization), meso-level (curriculum, learning design, educator/tutor), and micro-level (learner). Based on common data standards and open data schemas, rich datasets from virtual reality environments may enable the identification and validation of patterns within and across applications. At the meso-level, instructional designers may use analytics insights in order to improve VR learning artifacts or the sequencing of VR learning tasks as well as overall VR curricular design (Ifenthaler, et al., 2018). The micro-level focuses on VR learning processes of individual learners adapting to their current needs in the learning process and helping them to reflect and act on recommendations based on multiple analytics insights (Gašević et al., 2017).

An essential prerequisite of learning analytics benefits in VR, however, is the perspective of data and analytics involved (Chatti et al., 2021): (1) summative and descriptive; (2) formative and (near) real-time; (3) predictive and prescriptive. The summative and descriptive perspective provides detailed insights after completion of a VR learning phase, often compared against previously defined reference points or benchmarks. The formative and (near) real-time perspective uses ongoing data from VR environments to improve processes through direct interventions. The predictive and prescriptive perspectives are applied for forecasting the probability of outcomes to plan for future VR interventions, strategies, and actions.

Before utilizing learning analytics for VR, organizations are required to address data protection and privacy issues linked to learning analytics (Mutimukwe et al., 2022): They need to define who gets access to which data, where and how long will the data be stored, and which algorithms are implemented. Further, not all data from VR environments are relevant and equivalent for learning analytics and its analysis is critical for generating useful insights (Macfadyen & Dawson, 2012).

In moving forward to embrace the opportunities that could be provided by learning analytics for VR, the challenges that remain to be addressed must not be underestimated before organizations can use (semi-)automated analytics of complex competencies and understanding with confidence.

Author

• Dirk Ifenthaler, University of Mannheim
  Presenting Author