Using data anaytics to evaluate impact in online learning

The current disruption to education has impacted hugely not just on the lives of young people who will come to be known as the Covid generation, but on society as a whole.

Although the spotlight has been on educational technology, digital technology has reshaped human interaction well beyond education.

While other sectors of the economy including social media networks and retail platforms have highly sophisticated methods of gathering, analysing and benefiting from huge amounts of data, this has not yet impacted on learning analytics. Where it has been applied, it tends to be used for narrow applications or interventions. Our approach is to analyse learning networks to provide a more holistic view of one-to-one and group collaboration and investigate the impact of collaborative methodologies and materials. Improving collaborative learning is a desired learning objective but difficult to measure. With our approach, contributions will be mapped and measured.

Our overall objective is to analyse learning networks of teachers, resources and learners in order to recommend effective pedagogical interventions and measure their success. We address important questions in the current debate on education in a post Covid world:

What is the best model for blended learning?

What is the new role for the teacher?

Does a more connected learning network lead to better collaboration skills and student outcomes?

Can we replace the random selection of resources with pedagogical interventions that actually improve student performance?

Crucially, will improved analysis of the learning network improve results for students disproportionally disadvantaged at the moment?

The project involves a multi-disciplinary partnership between learning technologists, data scientists, game researchers, computer scientists and pedagogical innovators.

Klik2learn has an experienced team of programmers, pedagogical expertise and is the digital development partner for Scotland's largest FE college with access therefore to Moodle data.

All education organisations deploy some form of Learning Management System: either Moodle, or global brands such as Blackboard, Canvas, D2L and Google classroom. However, a deep understanding of these networked learning environments is lacking. We cannot clearly demonstrate impact on learning. As a spin-off, we could in future, offer a more objective predictor of grades than teacher estimates and moderation algorithms.

Our approach is founded on a strong grasp of current learning theories and a wide knowledge of the research on learning analytics dashboards. The initial focus will be on extracting relevant data and developing an analysis of the network: learners, teachers and resources -- broadly in line with the principles of socio-constructivist learning and communities of practice models.

The first pedagogical interventions will be focused on developing and measuring peer- to -peer collaborations including gamification techniques and simulations.

Further interventions will include investigating the impact of one-to-one and group tutoring.

The company has direct experience of building its own LMS: a Digital Learning Hub offering a networked environment of core products, a wide variety of additional learning resources and tools to facilitate tutor/learner interaction. This offers an ideal test bed for efficacy measurement using insights from larger networked environments.