Optimize CSCL Activities Based on a Data-Driven Approach

Abstract

Computer-supported collaborative learning (CSCL) has been widely used in the field of education. Computer-supported collaborative learning plays a very crucial role for improving learning performance, social interaction skills, problem-solving abilities, and knowledge building. However, most studies focus on implementing collaborative learning activities based on personal and subjective experiences. Previous studies seldom examine how to optimize collaborative learning activities based on a data-driven approach. This study aims to bridge this gap to propose how to optimize collaborative learning activities as well as evaluate the effectiveness of optimization strategies. Totally 72 junior school students participated this study and completed 7 collaborative learning tasks. For each collaborative learning task, two rounds of collaborative learning were implemented and recorded for analysis. The results indicated that the proposed 17 optimization strategies were very effective for improving the design quality of collaborative learning, the alignment between design and enactment, collaborative knowledge building level, and group products quality. The results and implications for teachers and practitioners are also discussed in depth.