This paper reports on a new movement in cognitive studies that focuses on understanding and promoting collaborative learning. First, learning goals are redefined and a theoretical explanation of how collaboration works to achieve such goals is provided. A laboratory study to test this theoretical framework is then described. The study was integrated within a two-year curriculum which teaches introductory cognitive science, and which uses technological tools to enhance teaching and learning processes and outcomes. Two classes are described: one using a technique called "the jigsaw method" to teach the construct of semantic memory, and the other using a more complex design called "the dynamic jigsaw" to facilitate sophomores learning how to integrate 24 research findings into a coherent view of cognitive science. The results to date are promising while at the same time stimulating new research questions about how college students may be helped in their acquisition of not just basic academic knowledge but also skills for selfdirected learning and collaborative work.Key words: learning sciences, collaborative learning, the jigsaw method, mechanisms of constructive interactionRecently, a new research field called "learning sciences" has been gaining popularity and redefining instructional practices worldwide. Learning sciences seek to establish theoretical understandings of what learning is, and to promote higher levels of learning. They iteratively refine learning-related theories and demonstrate the effectiveness of the devised methods through actual practices (Bransford, Brown, & Cocking, 1999;Bransford & Donovan, 2005;Miyake & Shirouzu, 2003).Studies of how people can effectively learn have a long tradition because human beings have constantly been required to re-structure old experiences and to accommodate them to new situations. However, in the current age, where rapid changes are the norm, the degree to which this requirement occurs has increased more than ever. Students, for example, are required not only to acquire routine knowledge but also to apply what they learn to new situations, outside of school and in a distant future. To meet such requirements, new learning studies need new research methods which differ from those used in laboratory-based, conventional learning studies (Brown, 1992;Collins, 1992).