Creating Data for Learning Analytics Ecosystems

Learning is a complex process that involves rich interactions between people, politics, places, and increasingly, technology. Using clickstream data to provide deep insights into learning requires care and a system-wide approach. We need learning analytics ecosystems. This position paper discusses the challenges faced by institutions as they seek to wrangle data from a portfolio of technology products and services to deliver rich learning experiences. It identifies a number of systemic failures that are leading many institutions to collect a large quantity of data that yields very few educationally relevant insights; from a lack of data policies to competition between vendors and standards. This position paper provides several high-level recommendations for three different sets of stakeholders to consider as they move towards a more mature implementation of LA: (i) Institutions procuring EdTech products; (ii) EdTech vendors seeking to deliver analytics products, and (iii) The LA community of practitioners. We also provide detailed evidence and examples in a series of appendices that are pitched at more technical members of your organisation as you seek to build a robust and useful learning analytics ecosystem.