There is a growing body of literature providing accounts of research that support the educational value of Computer Supported Collaborative Learning (CSCL). The results of a meta-analysis on CSCL reveal that using collaborative techniques with technology could increase high-level thinking skills, social interactions, critical reflective capabilities, and creativity (Lehtinen, Hakkarainen, Lipponen, Rahikainen, & Muukkonen, 2003; Smith, 2003; Warschauer, 1997).
Several empirical experiments offer some evidence that the well-known CSCL environments like Computer-Supported Intentional Learning (CSILE) and Belvedere have proved to be helpful for higher order cognitive processes and collaborative knowledge-building (Lamon, Reeve, & Scardamalia, 2001; Lehtinen et al., 2003). The learning efficiency of CSCL is further confirmed by various studies (Dewiyanti, Brand-Gruwel, Jochems, & Broers, 2007; Ewing & Miller, 2002; Gillies, 2004; Nachmias, Mioduser, Oren, & Ram, 2000; Resta & Laferriere, 2007; Shellens & Valcke, 2005) as, through this articulation process, old and new knowledge was integrated and new knowledge could be expanded to other applications. In these studies, CSCL appears to engage students to participate in in-depth inquiry over substantial periods of time and to provide socially distributed cognitive resources for comprehension monitoring and other meta-cognitive activities. This, in turn, allows students to become aware of their conceptual advancement, as well as of changes in their practices of inquiry.
Introducing a computer environment to collaborative learning can also improve the amount and quality of social interaction among learners and between educators and learners as these tools make the sequence of interaction events more visible for participants, opening better possibilities for mutual understanding. For example, a study conducted by Hakkinen, Jarvela, and Byman (2001) showed that the participants had mutual negotiations in their web-based communication and they discussed issues from a variety of different viewpoints. With the help of technology such as groupware, it has been possible to create interactive processes in which learners are consciously constructing new knowledge on an inter-subjective or social level (Kreijins & Kirschner, 2002).
In recent research, considerable attention has been paid to theoretical debate in the field of collaborative learning; however, little attention has been paid to issues of methodology and analysis methods to evaluate the quality of the collaborative learning community (De Wever, Schellens, Valcke, & Van Keer, 2006; Gunawardena, Lowe, & Anderson, 1997; Gunawardena, Carabajal, & Lowe, 2001; Strijbos, Martens, Prins, & Jochems, 2006) as they have been confronted with a range of problems. One of the problems of researching learning in CSCL environments is perhaps the realisation of the complexity of learning interactions being probed (De Laat, 2002; Lally, 2002; Lally & De Laat, 2002). The problem can be easily understood as it relates to the analytical tools used with the complicated procedures for content analysis. Content analysis is cumbersome and time consuming and the choice of coding categories is a complex issue in itself (Lally & De Laat, 2002). In this regard, extensive effort was made to devise an appropriate content analysis scheme for this study to illuminate interaction patterns and quality of the discourse of a CSCL community.
Activity Theory provides educators with a practical and holistic approach to the evaluation of a CSCL community (Hew & Cheung, 2003). To make the content analysis valid, there should be a concrete link between the analysis categories and the theoretical framework. Without a theoretical model of the collaborative learning process it is impossible to identify empirical indicators that will form the basis of a coding instrument as a standard against which to evaluate whether effective learning is occurring in the online discussions (De Wever et al. …