How well centrality measures capture student achievement in computer-supported collaborative learning? – A systematic review and meta-analysis
Abstract
Research has shown the value of social collaboration and the benefits it brings to learners. In this study, we investigate the worth of Social Network Analysis (SNA) in translating students’ interactions in computer-supported collaborative learning (CSCL) into proxy indicators of achievement. Previous research has tested the correlation between SNA centrality measures and achievement. Some results indicate a positive association, while others do not. To synthesize research efforts, investigate which measures are of value, and how strong of an association exists, this article presents a systematic review and meta-analysis of 19 studies that included 33 cohorts and 16 centrality measures. Achievement was operationalized in most of the reviewed studies as final course or task grade. All studies reported that one or more centrality measures had a positive and significant correlation with, or a potential for predicting, achievement. Every centrality measure in the reviewed sample has shown a positive correlation with achievement in at least one study. In all the reviewed studies, degree centralities correlated with achievement in terms of final course grades or other achievement measure with the highest magnitude. Eigenvector-based centralities (Eigenvector, PageRank, hub, and authority values) were also positively correlated and mostly statistically significant in all the reviewed studies. These findings emphasize the robustness and reliability of degree- and eigenvector-based centrality measures in understanding students’ interactions in relation to achievement. In contrast, betweenness and closeness centralities have shown mixed or weak correlations with achievement. Taken together, our findings support the use of centrality measures as valid proxy indicators of academic achievement and their utility for monitoring interactions in collaborative learning settings. © 2022 The Authors
Affiliations
University of eastern Finland, School of Computing, Joensuu, Yliopistokatu 2, fi, Joensuu, 80100, Finland; University of eastern Finland, Philosophical Faculty, School of Applied Educational Science and Teacher Education, Yliopistokatu 2, fi, Joensuu, 80100, Finland; Departamento de Sistemas Informáticos, Universidad Politécnica de Madrid, ETSI Sistemas Informáticos, c/ Alan Turing, s/n, Madrid, 28031, Spain