Collaborative knowledge construction: examples of distributed cognitive processing Michael Tscholl John Dowell Department of Computer Science University College London, UK Abstract In discussions centered on jointly solving a problem or case, significant construction of new knowledge can occur. Several theoretical frameworks have been invoked to explain the productivity of dialogues, but questions about the relative or unique impact on learning of the multiple facets of dialogues remain. We present an analytical approach that studies small‐group discussions from the perspective of joint cognitive processing of knowledge and information. We illustrate our approach through a microanalysis of two discussions that were held in a real‐world educational setting. We show that knowledge construction can overlap significantly with critical argumentation, but may occur even in its absence. On the basis of these findings we propose a refined definition of co‐construction, and a view of the inter‐relations between interaction and co‐ construction. We discuss the implications of our findings for the analysis and evaluation of differences in knowledge co‐construction in different environments. Keywords Collaborative Learning Interaction Analysis Synchronous Communication Cognitive Perspective Case‐based Learning INTRODUCTION In discussions centered on jointly solving a problem or carrying out a task, significant construction of new knowledge can occur. Each problem represents a new challenge for a team and requires the use and adaptation of existing knowledge. The productivity and creativity of groups involved in joint problem solving is well known, although the processes by which groups jointly construct knowledge have only recently been studied in detail. Beyond the relevance of research studying these processes for social learning theories, more direct interest in this research is in its potential output for the evaluation of computer systems supporting virtual learning groups. In the last decade, substantial research has addressed issues such as how to support distributed groups, how to analyse distributed group processes and how distributed processing of knowledge relates to learning gains. This chapter contributes to this literature by presenting an analytical approach by which the joint knowledge construction processes as well as the potential learning gains from such processes are described and modeled. We rely on a microanalysis of dialogues that focuses on relations