A Step toward Characterizing Student Collaboration in Online Knowledge Building Environments with Machine Learning

Abstract

Existing research has substantial progress in uncovering outcomes of collaborative learning in recent years, but more attention can be directed towards the better understanding of collaborative learning processes via quantitative frameworks and methods. Through the use of knowledge building as a collaborative learning pedagogical approach, it is possible for researchers to glean deeper insights into aspects of students’ collaboration within authentic learning environments. In this paper, the multimodal approach of data collection and analysis was conducted with a proposed conceptual analytical framework that can characterize constructs of collaborative activities in a knowledge building classroom using machine learning methods. The application in a pilot is discussed along with how this conceptual development can offer a summary of new insights into students’ individual and group collaborative trajectories during learning tasks.