Come sfruttare gli Educational Data ? Un inquadramento di usi e metodologie di analisi

This article provides an overview of the most relevant statistical methods used to explore educational data and to derive key analytical insights from them. We describe two broad categories of statistical methods : those pertaining to Educational Data Mining and Learning Analytics. The former category gathers the methods aimed at uncovering hidden patterns from educational data, with the aim to improve and facilitate the task of all the actors involved in the education process, like students, instructors, and administrators. The latter category gathers the methods specifically oriented to measure how much student learn and how. A detailed explanation of the most popular statistical methods into both categories is offered to the reader, as well as an analysis of their similarities and differences. We conclude that Learning Analytics is typically more concerned with pedagogical aspects involving human judgment, while Educational Data Mining is more oriented to identify in automated way optimal practices to improve the outcomes of the learning process from student feedback.