Effectiveness of the cooperative base groups technique in facilitating cooperative learning in small WhatsApp groups for first-year computer science students: a multi-level social network analysis
Christina Johanna van Staden, Liezel Nel
Discover Education
Network design
Nodes
The students allocated to each of nine cooperative base groups, together with the lecturer, assistant, and a constructed shared-group information node used in that group's network model.
Ties
A directed flow of academic information, advice, support, expertise, leadership, or another learning resource from a provider to a receiver in WhatsApp; repeated uses increased the tie weight.
Methods
Nine directed weighted edge lists visualized with Fruchterman-Reingold layouts in Gephi; macro density, distance, degree, and component measures; meso modularity and community detection; micro clustering and centralities; personal development network inspection; and Pearson correlations with final marks.
Source: Discover Education
Reviewed summary
The lecturer randomly allocated 79 students in one South African computer science module to nine cooperative base groups of eight or nine members for semester-long WhatsApp activities. Students were expected to meet weekly, complete routine tasks, encourage submissions, and provide academic and personal support. Eleven allocated students never joined or later left their groups, and participation patterns differed sharply across the nine bounded networks.
The researchers reconstructed a directed, weighted development network for each group from message exchanges. None of the nine networks contained every possible relationship, six were fragmented, and communities or strongly connected components appeared within the groups. Ten students had no recorded development relationship. Personal-network size correlated weakly with final marks overall, but central students were not consistently the highest achievers, so the patterns do not demonstrate that network position or the group technique caused academic performance.
SNA
How SNA was used
Each WhatsApp group was analyzed as a separate whole development network at three levels. Macro analysis used degree, weighted degree, density, average path length, diameter, and components; meso analysis examined connected groups and modularity-based communities; micro analysis used clustering, degree, eigenvector, betweenness, closeness, and personal development networks. Pearson correlations compared personal-network size with final marks as a separate, non-causal triangulation.
Source
Discover Education, 4, Article 552
DOI: 10.1007/s44217-025-00972-y
