Smart learner support through semi-automatic feedback

Abstract

Learning tools that produce automated feedback are becoming commodity, from multiple-choice questions to intelligent tutoring systems, and from direct manipulations to exploratory environments. In this paper, we argue how such learning tools can become smart by applying the semi-automatic feedback paradigm where the teacher complements the feedback capabilities of the learning tool. The approach employs analytics as a central awareness mechanism for teacher to provide guidance in a way that is most relevant to the past usage of the learning tool, including what it provided as feedback. The SMALA approach we describe is realized as an open-source software which has been evaluated in a number of undergraduate studies, leveraging the default learning management system’s architecture of the universities. This software delivers visualizations of the activities at each level of interaction (the group of all users, the group of users in a classroom, the individual learner). The different levels support the teacher in adjusting his or her strategy and respond to individual requests.