An Observation Framework for Recognising Learning Evidence in 3D Collaborative Virtual Environments

Immersive environments such as 3D virtual spaces enable collaborative learning and facilitate better connections between students, virtually. Learners do acquire new knowledge or skills while practising collaborative activities in such spaces. However, recognising evidence of learning to assess students is a critical issue which must be considered when organising learning activities in virtual environments. Although there is extensive coverage in the empirical literature regarding assessing learning in real-world classrooms, there is a lack of research focused on identifying learning evidence and assessing students who are performing educational activities within virtual worlds. This thesis aims to fill this research gap, exploit the affordances of immersive environments, and investigate appropriate methods for identifying users’ performance within these. This research proposes a computational framework, and a number of virtual observation models, for classifying learning evidence in immersive environments – and then maps all these elements to an appropriate learning design. In order to implement the computational framework required, the research includes the construction of a proof-of-concept prototype. The prototype employs virtual observation components and applies fuzzy logic and multi-agents approaches in order to assess students’ performance in real-time; this is from a number of different perspectives and based on multiple pedagogical frameworks. The present study also goes on to evaluate the research framework proposed by putting together a large number of educational sessions which are then carried out in a virtual world. These evaluation sessions involve both student and expert participants collaborating together to validate the model used. Subsequently, the thesis discusses the findings from the experimental sessions and their broader significance for the research area. Overall, the results strongly supported the effectiveness and usefulness of using the proposed virtual observation method when assessing collaborative students performing within immersive environments.