Learning analytics in practice: Providing E-learning researchers and practitioners with activity data

Abstract

In this chapter, we describe a practical solution for providing all researchers and practitioners in an online university with a unified learning analytics database (LA-DB) containing evidence-based activity data. Our goal is to seamlessly capture all relevant data generated within a virtual learning environment, using a very simple learning record store containing only a few tables, trying to overcome the typical problems in such a huge and complex scenario, namely data fragmentation, duplicity, inconsistencies, and lack of standardization across different data sources currently used by the university, without interfering with current information systems and procedures. In order to do so, some technological and organizational changes to promote a “data culture” within the institution have been considered. The system, implemented entirely using cloud services, allows researchers and practitioners to pose and answer questions using a simple activity-driven data model, combining data from three different levels of analysis, ranging from session-based (short-term) to institutional (long-term). Available data includes navigation, interaction, communication, and assessment, as well as high-level indicators that aggregate and summarize learner activity. Finally, we also present some preliminary actions taken for fighting early dropout as an institutional project using the proposed infrastructure and gathered data.