Abstract:
Computer-Supported Collaborative Learning (CSCL) emerges as an instrument
of learning and training that can encourage the social nature of learning by
adopting range of computer-mediated communication tools and pedagogical
methods. These tools are used to facilitate the learning and instructional communication among students and learners in small groups. In this process, group
formation becomes painstaking and challenging task. Various factors are involved affecting group formation that includes; personal characteristics, social,
cultural, psychological and cognitive diversity. Although, this issue was addressed in various research studies yet an optimal solution for dynamic group
formation is not discussed evidently. In dynamic groups, students work or collaborate on a short term tasks in a group that change frequently based on the
performance of students. In this research study we have proposed a method for
dynamic group formation and the impact of dynamic group formation in collaborative learning among peers is demonstrated by conducting an experiment.
This experiment is conducted in two phases. In first phase of the experiment,
learning styles are assigned to the students and their knowledge level is calculated. Whereas, in second phase of the experiment, the impact of dynamic
grouping on collaborative learning of students is determined. Further, two algorithms are proposed, first is used for determining initial number of clusters
and second algorithm is used for dynamic grouping after the completion of each
permutation. Results of our experiment shows a positive impact of dynamic
grouping on learning of students as the performance of collaborative learning
among peers is better than individual performances.