Deconstructing online social learning: network analysis of the creation, consumption and organization types of interactions. Issue 1 (1st January 2020)

• Record Type: Journal Article
• Authors: Kent, Carmel; Rechavi, Amit
• Is Part Of: International journal of research & method in education. Volume 43:Issue 1(2020)
• Page Start: 16
• Page End: 37
• Publication Date: 2020-01-01
• Subjects: Social learning analytics -- social network analysis -- online communities -- collective learning; Education -- Research -- Periodicals; Education -- Periodicals; 370.705
• DOI: 10.1080/1743727X.2018.1524867
• Languages: English
• ISSNs: 1743-727X

Abstract

ABSTRACT: Educational research suggests that interactivity is one of the most important tools for learning. This paper analyses the learning process in online communities by examining three types of interactions among learners: (1) interactions involving the active contribution of content ('digitally speaking'); (2) interactions involving the consumption of content ('digitally listening'); and (3) interactions involving the organization of digital content. We present a network analysis framework for online discussions entailing both static and dynamic analyses of these types of interactions. We build three sub-networks deconstructed from online discussions in four different learning communities. The suggested framework for network analysis follows the paradigm of collaborative learning, and thus, it analyses the dynamics of the collective learning, rather than individual learning. We use topological indicators, such as the distance between learners, reciprocity, and the individual's influence on the collective, to explore the networks between learners and to investigate their evolution during one semester. Interactional patterns in the sub-networks are different and sometimes contradictory, and the relationships found among speaking, listening and organizing digital content might be rooted in different learning designs. Our findings suggest that social learning assessment should address different interaction types separately.