| Sumario: | This article examines the diverse methodological approaches used in seven studies on collaborative learning, highlighting how different theoretical frameworks shape data collection, coding, and analysis methods. It emphasizes the predominance of multimodal data sources—including discourse, behavioral traces, self-reports, and artifacts—and the use of both qualitative coding and advanced computational techniques such as network analysis, sequence mining, and machine learning. The article critically reflects on the interpretive nature of coding practices, noting their theory-laden and selective character, and advocates for incorporating participants’ perspectives and negative case analysis to enhance validity. Central to the discussion is a call for shifting from analyzing networks of words or codes toward modeling networks of inferences—structured relations among claims and reasons—to better capture the meaning-making and reasoning processes underlying collaborative learning. This inferential perspective aims to complement existing methods by representing how shared understanding and regulation emerge through evolving argumentative structures across interactions and contexts.
|
|---|