Socio-spatial Learning Analytics

Capturing socio-spatial aspects of learning is vital for classroom orchestration and collaborative learning in complex spaces. Traditional methods like surveys and observations have limitations. This thesis addresses gaps in socio-spatial learning analytics: 1) the need for theory-grounded methodologies, 2) ensuring ecological and statistical validity, and 3) considering learning design. We developed methodologies based on traces from two large-scale studies in authentic learning settings. We also created a framework for socio-spatial learning analytics, incorporating theory, learning design, and evidence-based methods. The findings impact future analytics, emphasising collaborative learning's socio-spatial facets, instructional design effects, and ecological studies. Ethical considerations are also discussed.