Analytics

In the past decade, interest has emerged in the large amounts of data generated by the increasing use of technology in education and what we can learn from it. Communities of researchers have arisen around the fields of learning analytics and educational data mining to look for patterns in various types of digital activity. Learning Management Systems can track every click a student makes from the time they enroll to the time they leave an educational program using it. Artificial Intelligence (AI) techniques have been developed to analyze the specific online interactions and adapt learning experiences based on this.

Learning analytics research is often purely quantitative in approach and does not provide the full context surrounding student behavior. We can show “what” is happening (e.g. large drop-offs in participation in MOOCs), but not “why” it is happening. The “why” is important if we are to understand the motivations of students and how we can best design interventions to assist them. Artificial intelligence (AI) approaches can analyze how effectively and efficiently students move through a course, but not how satisfied they are with the experience. There is a link to research in social networks, which may lead to a richer set of interventions to support student success. Data science should be complemented with the qualitative methods of social science to help get to deeper issues involved in student interactions with educational experiences.

ASU’s online learning support division EdPlus has developed an Action Lab to set up the data collection infrastructure and coordinate the research expertise to make progress in this relatively new area. It is providing a unique angle to this work by integrating social scientists and data scientists with the goal of using data not only to improve the persistence and course completion of students, but to understand their motivations and how they can be best supported in long-term achievement.

Types of analytics