Visualising Individual Profiles and Grouping Conditions in Collaborative Learning Activities

Abstract

Collaborative learning has been shown to be conducive to better and deeper learning for particular tasks, but is dependent on a number of factors, including how students are grouped together. We are interested in finding out whether data captured from students working individually and/or collaboratively can reveal useful information about the impact of the grouping conditions on learning. We explore whether these findings can be detected early on (possibly, before students start working in groups). If such information can be reliably captured, then it could be used to drive group formation dynamically and at a large scale. This paper presents our initial visual exploration with two case studies: one from a first-year programming course (N = 372) where students alternately worked individually and in pairs; and another (N = 60) from a concept mapping environment where students first worked individually and then in groups.