Learning sequences and events in collaborative data-driven learning

• Jarvis Looi ^[1]
  1. [1] CNRS-Université de Lorraine
• Localización: EuroCALL 2025. Advancing CALL: New research agendas / Yazdan Choubsaz (dir. congr.), Paz Díez Arcón (dir. congr.), Ana Gimeno Sanz (dir. congr.), 2025, ISBN 978-84-1396-326-6, págs. 333-341
• Idioma: inglés
• Enlaces
  • Texto Completo Libro
• Resumen

  • This paper explores how learning unfolds in collaborative data-driven learning (DDL), focusing on the observable behaviours and interaction patterns that shape inductive learning. Drawing on sociocultural theory, constructivism, and the noticing hypothesis, the study examines field recordings and learner worksheets from six DDL activities carried out by 61 multilingual learners of French. Using grounded analysis, twelve learning events were identified and grouped into seven higher-level learning sequences: observation, rule extraction, simultaneous observation and extraction, rule application, rule testing, blockage, and teacher support. These sequences emerged not as a fixed progression but as recursive, flexible cycles shaped by task demands and group dynamics. Scaffolding and languaging were central across all phases, while noticing alone did not consistently lead to learning without group consensus and further induction. The findings challenge the linearity of traditional DDL frameworks such as the 4 Is model and suggest that learning processes are socially mediated, cyclical, and iterative. This study contributes to DDL research by linking observed learning behaviours to SLA theories and highlights implications for the design of collaborative and inductive learning activities.