Multimodal behavioral profiles of dyads' collaborative learning are identifiable by integrating video, audio, and log data, with emergent processes more explanatory than individual behaviors¶

Claim · All claims
Evidence · 1 study · theoretical r2 · q2
Currency · newest study 2021 · not yet checked against its sources · not reviewed by a person · overdue for a recheck since 2024: tool and model accuracy changes quickly

Subclaims¶

q2 i? The editors report that integrating video, audio, and log data identified three multimodal behavioral profile clusters for dyads (Expressive Explorers, Calm Tinkerers, Silent Wanderers), and that emergent underlying processes are more explanatory than individual behaviors. → Carolyn Rosé 2021

Evidence¶

Carolyn Rosé 2021¶

Carolyn Rosé, Sanna Järvelä. (2021). A theory-driven reflection on context-aware support for collaborative discussions in light of analytics, affordances, and platforms. International Journal of Computer-Supported Collaborative Learning. https://doi.org/10.1007/s11412-021-09360-8

q2 · i? · theoretical · r2

Editorial account (the editors describe Nasir et al.'s multimodal study of dyads on an interactive tabletop with a robot mediator): affect is included to connect behavior and learning, and clusters named Expressive Explorers, Calm Tinkerers, and Silent Wanderers; "underlying intents, strategies" are the explanatory level.

"Emergent processes signify the existence of underlying intents, strategies and emergent processes, which on the one hand might be harder to pinpoint, but on the other hand are more explanatory."

Discussion¶

Related Claims¶

• Certain multimodal data combinations improve predictive model performance, with audio plus eye-tracking data most effective in one K–8 study
• Calm Tinkerers display significantly lower negative valence and arousal than Silent Wanderers, while Expressive Explorers show high arousal and negative valence similar to Silent Wanderers — a narrower finding that bears on this claim
• Among 32 dyads in an open-ended collaborative tabletop task, three behavioral clusters split into two learning (gainer) profiles and one non-learning profile despite similar task performance — possibly the same claim (merge candidate)
• Latent profile analysis of 2015 PISA data identifies four collaborative problem-solver profiles among Chinese 15-year-old students — related
• Speech activity and speech overlap distinguish gainer dyads from non-gainer Silent Wanderers across both gainer profiles — a narrower finding that bears on this claim
• Multimodal data can reverse the interpretation of silence in collaborative problem-solving — related
• The two gainer profiles differ in problem-solving strategy: Expressive Explorers enact global exploration with reflection on past solutions, Calm Tinkerers enact local exploration with in-the-moment correction — a narrower finding that bears on this claim
• Certain conditions of complex emergence can be deliberately tinkered with in educational settings — related