Beyond Small Groups: New Opportunities for Research in Computer-Supported Collective Learning

CSCL research has focused on understanding and designing collaborative learning in diverse settings and configurations with support of computers. Within this research, however, most efforts have concentrated on studying small group configurations and thus examined what we would like to call ‘collaborative’ learning (i.e., the abilities needed to participate and support collaborations of typically two to five people). Much less emphasis has been placed on studying massive communities and participation in large groups prominent in today’s social networking sites and online gaming cultures that would shift the focus to ‘collective’ learning (i.e., the abilities needed to participate and support collaborations in massive groups). In this paper, we identify key dimensions of collective learning, present observations of online and local participation in one open-source Web 2.0 community with over 630,000 members, called Scratch (scratch.mit.edu), and outline a research agenda for computer-supported collective learning. Introduction Research in CSCL, by its very name, has focused on understanding various dimensions of group work such as productivity of different group arrangements (Engelmann & Hesse, 2010), design of scaffolds (van der Pol, Admiraal, & Simons, 2006), argumentation practices (Scheuer, Loll, Pinkwart, & McLaren, 2010) and interactions between online and offline collaborations (Birchfield & Megowan-Romanowicz, 2009). With some exceptions (e.g., Fields & Kafai, 2009; Guzdial & Rick, 2006), there is one assumption about collaboration underpinning many of these efforts, which is the idea that collaboration happens in small groups, often of dyads and triads, as they engage in computer-supported collaborative tasks. For the most part, the focus of this research has been to understand and develop what we will refer to as collaborative learning because it emphasizes the abilities to participate in small groups whether online or offline or in combinations thereof. Recent developments, however, suggest new forms of collaborations are developing in online communities (e.g., Boyd, 2008; Shirky, 2008). One striking feature of these communities is their size and collaboration that can take place among hundreds, if not thousands, of members. Consider the millions of contributions to entries in the Wikipedia or to programs in Linux (Benkler, 2006), the interactions of members of fan fiction sites where thousands of writers create new stories and participants provide constructive feedback (Black, 2006), or the participation in guilds in multiplayer online role-playing game communities with millions of players (Gee, 2003). We have chosen to call this type of collaboration collective learning because it emphasizes the abilities to participate and perform in collectives and thus might be different from participation in small groups. Our concept of collective learning is inspired by recent research in different communities: the work of social scientist Pierre Levy on collective intelligence (1997) examining the potential of intellectual contributions from large groups; the work of gaming researcher Jane McGonigal (2008) observing selforganized coordination among players in collective gaming; and the work of media scholar, Henry Jenkins (2006) studying participatory culture in networked communities. Taken together, this body of work converges, helping us to recognize that large-scale communities can promote new forms of opportunities, as well as challenges, for learning together. As increasing numbers of learning communities in K-12 and higher education move online, so too grows our need for understanding how to engage large groups of learners effectively in these networked communities. The starting point for our investigation on collective learning is a simple question: Does the size of the group matter? We realize that there will be no simple answer because massive communities such as social networking sites, virtual worlds, and multiplayer online role-playing games are each organized around different purposes and practices and thus participants have different incentives for joining and collaborating with each other (see also Hung, Lim, Chen, & Koh, 2008). We begin with a review of what research has identified as key features of collaborative learning in small and large groups. We then draw upon research and our knowledge of one such large-scale networked community, called Scratch (scratch.mit.edu), to describe and articulate the different dimensions of collective learning. In our discussion, we outline the emerging challenges and opportunities for research in computer supported collective learning. Background The work in CSCL draws upon hundreds, if not thousands, of research studies that have investigated various aspects of collaboration, including the nature of various group arrangements such as reciprocal teaching or CSCL 2011 Proceedings Volume I: Long Papers