A priori knowledge in learning analytics

Abstract

Learning Analytics (LA) can be data driven: the process is oriented essentially by data and not according to a theoretical background. In this case, results can be, sometimes, not exploitable. This is the reason why some LA processes are theory driven: based on A Priori Knowledge (APK), on a theoretical background. Here, we investigate the relationship between APK and LA. We propose a “2-level framework” that considers LA as a level 2 learning process and includes five components: stakeholders, goals, data, technical approaches and feedbacks. Based on this framework, a sample of LA related works is analyzed to exhibit how such works relate LA with APK. We show that most of the time the APK used for LA is the learning theory sustaining the student’s learning. However, it can be otherwise and, according to the goal of LA, it is sometimes fruitful to use another theory.