Collaborative learning with distributed information

In this project, we address a common design dilemma that instructors face when they want to structure collaborative learning: should they design for knowledge interdependence by distributing core learning content between students, as in a classical jigsaw script? Or should they design for common ground by emphasizing shared knowledge as a basis for productive collaboration?

We argue that weak, rather than strong knowledge interdependence provides a sensible solution to this dilemma (Deiglmayr & Schalk, 2013). In particular, instructors should ensure that learners have sufficient common ground regarding core domain principles; that is, core learning content should not be distributed between learners. Instead, each learner might study the same core concepts embedded in a different context, applied to a different case, or introduced in a different representational format. This “weak” form of knowledge interdependence ensures that learners’ complementary knowledge renders collaboration relevant and fruitful, while at the same time ensuring that knowledge about core domain principles is shared among learners from the start.

In our studies, we work with extensive analyses of process data (self-explanations, chat, log files). We base these analyses on the recently proposed ICAP-framework (Chi & Wylie, 2014), which posits interactive (i.e. co-constructive) activities of engaging with learning content to be the most beneficial for individual learning. As part of our project, we also put this claim to a stringent test. For example, we could show that interactive learning activities could, indeed, partially explain the beneficial effects of weak over strong knowledge interdependence on individual learning (Deiglmayr & Schalk, 2015).