Networked Learning Analytics: A Theoretically Informed Methodology for Analytics of Collaborative Learning

Abstract

Online social learning is a prevalent pedagogical tool, enabling learners across all ages and cultures to learn together. Educators, policy-makers, and international organizations such as the Organization for Economic Cooperation and Development (OECD) stress the need to assess collaborative learning systematically. However, the systematic assessment of large online groups’ collaboration is still in its infancy. In this chapter, we suggest perceiving social learning through the lens of interaction networks between learners and content. Based on well-accepted learning theories, we demonstrate the harnessing of digital traces of online discussions to the assessment of social learning, at both the individual and the group levels. Practically, our contribution is to suggest a network analysis point of view for the assessment of the performance and design of learning communities. Our proposed methodology can be used by instructors to open-up the black box of collaborative learning, to be able to equip learners with twenty-first-century skill-set.

Appendices

Appendix A: Topological Parameters—Brief Index

While this chapter cannot elucidate all SNA concepts, we bring here the necessary SNA definitions. We refer the reader to a canonical text about networks and their topological parameters in Wasserman and Faust (1994) and Easley and Kleinberg (2010). For the general note on the importance of SNA concepts, we refer the readers to Rafaeli (2017).

• Authority and Hub Degree: Authority and hub values are defined in terms of one another in a mutual recursion. An authority degree of a node is computed as the sum of the scaled hub values that point to that node. A hub value is the sum of the scaled authority values of the nodes to which it points. A node has a high authority degree if numerous hubs point to it and a high hub degree if it points to numerous authorities. As Kleinberg (1999) puts it: “…, a review paper may refer to other authoritative sources: it is important because it tells us where to find trustworthy information. Thus, there are … two types of central nodes: authorities, that contain reliable information on the topic of interest, and hubs, that tell us where to find authoritative information.”

• Betweenness Centrality: For each node, the betweenness centrality is the number of shortest paths between any two nodes that pass through this node. High betweenness means that a node is connecting many pairs of nodes that would not be connected without it.

• Clustering Coefficient: The network’s clustering coefficient measures the degree to which nodes in a network cluster together. In social networks, in which nodes (people) tend to create groups with a relatively high density of links within them, a high clustering coefficient is expected. For a single node, it is the ratio of network links connecting a node’s neighbors to each other to achieve the maximum (potential) number of links.

• Burt’s Effective Network: The number of nodes a given node can reach in the network. It measures the overall reach of a node in a network.

• Clique: A group of nodes in which each individual node is directly linked with every other individual. For example, Alice follows Bob and Charlie, Bob follows Charlie, and Charlie follows Alice. In this chapter, cliques need not be reciprocally connected.

• Eigenvector Centrality: The connections of a node to high-scoring nodes are assumed to contribute more to the score of the node in question than would an equal number of connections to low-scoring nodes. The rationale behind this centrality score is that if Alice is a friend of Bob and Bob knows everybody, Alice can benefit from Bob’s connections. Eigenvector centrality assigns relative scores to all nodes in the network to measure the influence of each.

• Exclusivity: The degree to which the node is the only node that connects to other (less popular) nodes. Exclusivity indicates that the node in question has access to unique nodes (people or resources).

• In-Degree: The number of incoming links a particular node receives.

• Out-Degree: The number of outgoing links of one node to other nodes.

• Total Degree: The sum of in- and out-degrees.

Appendix B: Topological Change in the Network (Fig. 9.9)

From right to left: first week, fourth week, and eighth week. The brown nodes are posts, and the central one is Post 1, the root of the discussion (including its title). The brown nodes in the second tier are Posts 2 to 10 that constitute the skeletal nodes, including the course’s subject. The green nodes depict the learners. Nodes with the highest betweenness centrality appear in the center of the network

Fig. 9.9

Network evolution

Appendix C: Daily Changes in Diameter and APL (Fig. 9.10)

Fig. 9.10

Daily changes in the network’s diameter (red) and APL (blue)

Appendix D: Ligilo Screenshots (Fig. 9.11)

Fig. 9.11

A basic view of Ligilo

Posts at the left side of the screen are connected by blue tagged relations to posts on the right.

Fig. 9.12

Bird’s eye view of Ligiloposts

Fig. 9.13

Tagging relations in Ligilo