Model-based clustering of tied durations identifies four weekly regulation patterns in PBL group discussions¶

Claim · All claims
Evidence · 1 study · associational r? · q2
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i? Four weekly regulation patterns emerged from clustering groups' temporal interaction data. → Mohammed Saqr 2026

Evidence¶

Mohammed Saqr 2026¶

Mohammed Saqr, Sonsoles López-Pernas, Tiina Törmänen. (2026). A temporal network approach to reveal the longitudinal dynamics of CSCL group regulation and productive collaboration. Intern. J. Comput.-Support. Collab. Learn. https://doi.org/10.1007/s11412-025-09464-5

q2 · i? · associational · r?

Model-based clustering (Gaussian mixture models via MoEClust) of each group's weekly tied durations, evaluated across "nine regulation patterns ranging from two to ten". The "solution with four regulation patterns solution was optimal" per the article's fit criteria; Table 2 reports BIC of −6070.49 for the four-cluster solution.

"Using a combination of these criteria, we estimated nine regulation patterns ranging from two to ten. We found that the solution with four regulation patterns solution was optimal"

Discussion¶

Related Claims¶

• Most PBL groups settled into stable regulatory patterns early in the course, with transitions usually between similar patterns
• Optimal clustering of five-channel CPS sequences detected three types of collaborative patterns among graduate student groups
• Students using interactive socioemotional or regulated constructive patterns were more productive cognitively and in knowledge productivity — related
• Gaussian Mixture Modeling identifies two latent clusters of delayed start behavior, outperforming three-cluster and unimodal models — related
• Trace data can capture the temporal progress of regulation as learners engage in learning tasks over time — related