Scaling up learning analytics in blended learning scenarios
- Scaling up Learning Analytics in Blended Learning Szenarien
Lukarov, Vlatko; Schroeder, Ulrik (Thesis advisor); Verbert, Katrien (Thesis advisor)
Dissertation / PhD Thesis
Dissertation, RWTH Aachen University, 2019
In recent years, there is a prominent claim that learning analytics is a key transformative action that will radically transform education and its processes. This field draws its roots and methods from data analysis, statistics, data mining, business intelligence, computer science, and educational research and learning. Extensive research has been done to develop tools, prototypes, and analyze educational data to improve and innovate education, and this has advanced the research field of learning analytics. However, this has created a widening gap between what could the role of learning analytics be in education, and what learning analytics is actually doing in education. The research evidence shows that the use of learning analytics to improve learning and education is still in its infancy, and there is a lack of practical examples and implementations on scale and practical approaches of how to provide learning analytics services in education and put them into practice. This dissertation focuses on the practical problem of scaling up learning analytics services in blended learning scenarios in a higher education institution in Germany. This dissertation presents the solution as a set of key principles for scaling up learning analytics in blended learning scenarios in higher education. These focus on five aspects: collecting correct requirements for the different stakeholder groups, preparing the legal and technical foundations of the higher education institution, continuously develop and improve the learning analytics services, and continuously evaluate the learning analytics services. These aspects were comprehensively investigated and realized by applying design-based research methods, software engineering methods, and evaluation methods from the Human-Computer Interaction field and the behavioral and cognitive sciences. The results and contributions from this research work are a verified end-to-end process for scaling up learning analytics in a higher education institution in Germany, a comprehensive set of requirements for the stakeholder groups, a categorized and comprehensive set of learning analytics indicators from the research and practitioners community, a sustainable learning analytics infrastructure with optimized analytics engine for scalability and performance and high fidelity prototypes, and a validated method for longitudinal studies for learning analytics impact evaluation. The basic idea is to enable another development team, or institutions from Germany to take this dissertation, its guidelines and results and use it to scale up learning analytics services at their higher education institutions.