Toward an inclusive understanding of collaborative learning using MMLA: Exploring multimodal participation dynamics by gender and linguistic diversity

Background Sociocultural learning theories highlight that knowledge is socially constructed through interactions among individuals, tools, and the environment. To foster a more inclusive understanding of collaborative learning, it is important to explore how learners from diverse backgrounds participate. However, capturing and analyzing multimodal participation in real-world settings remains a challenge. Aims This study examines how gender and linguistic background differences shape multimodal participation dynamics in collaborative learning groups within a real-world, out-of-school Computer-Supported Collaborative Learning (CSCL) environment. Sample Participants included 33 middle school students with diverse cultural and linguistic backgrounds, participating in a bilingual, integrated computer science and mathematics program. Five student groups varied in gender composition (female-dominant vs. male-dominant) and linguistic background (multilingual vs. English-dominant). Methods We employ an innovative MMLA approach that integrates an Artificial Intelligence-based video analysis system and multimodal data modeling using Epistemic Network Analysis, layered with qualitative video-based interaction analysis, to analyze multimodal participation dynamics. Results While individual-level differences based on gender or linguistic diversity were not significant, the results revealed notable differences in multimodal participation dynamics at the group level. The distinct patterns exhibited by groups with dominant gender and language backgrounds illuminate the varied pathways through which roles, tasks, and ideas are negotiated and contributions are made. Conclusions This study highlights the need to value and attend to multimodal participation dynamics, particularly by recognizing nonverbal modes of participation. We can take meaningful steps toward creating inclusive CSCL environments by leveraging the strengths of diverse learners, embracing a multimodal perspective, and adopting innovative methods to capture the complexity of CSCL.