Automatic student group generation for collaborative activities based on the zone of proximal development

Abstract

The Zone of Proximal Development (ZPD) theorized by Lev Vygotskij can be considered as one of the most interesting insights on the value of collaborative learning: it specifies the cognitive distance reachable by a student in learning activities, given that some support by teachers, and peers, is available. Group handling is certainly one of the main aspects to be reproduced in an online learning system. This paper aims to offer a twofold contribution. First, with a proposal for the evaluation of the “pedagogical quality” of the partition of a class in groups, over a specific activity, in terms of internal homogeneity, and external balance of expected effort. Second, with policies to automatically, or semi-automatically, make groups according to the cognitive characteristics of single students. An evaluation of group composition is also provided according to a computation of the group ZPD, which is estimated basing on a notion of Daring Threshold, shaping the maximum viable effort.