Intuitive moderation styles and beliefs of teachers in CSCL-based argumentation

CSCL learning environments provide new contexts for discussions and are thought to provide new opportunities for learning. At the same time, such environments often do not provide guidance on how to act during the discussion. The purpose of this paper is to initiate research on moderation in synchronous discussions in a CSCL environment. The first study contrasts teachers' beliefs on good discussions and good moderation pertaining to face-to-face discussions with those pertaining to synchronous, CSCL-mediated discussions. The second study focuses on the strategies teachers intuitively enact in synchronous discussions. The challenge of moderation in CSCL-based argumentation Learning processes in classrooms are influenced by the quality of discussions in which the teacher engages with the students, and particularly the teacher's moderation practices (Mercer, 1995; Chi, Siler, Jeong, Yamauchi & Hausmann, 2001; Wegerif, 1996; Atzmon, Hershkowitz & Schwarz, 2006). Oral discussions are generally teachercentered. The teacher raises questions, directs answers, stresses important issues, selects speakers, summarizes and presents new points into the conversation. CMC tools such as CSCL learning environments provide new contexts for discussions and are thought to provide new opportunities for learning: they enable learners to follow the developing interactions among others, to mutually examine the extent and nature of their own involvement in the process, and at the same time and to create awareness of the processes of self-thinking (Lave, 1991). At the same time, such environments rarely provide guidance or direction concerning how to act during the discussion (Soller, 2001). Types of discussions in classrooms are varied. Among these types, collective argumentation is particularly important for learning purposes. The object of argumentation, the elaboration of arguments, has been recognized as central in knowledge acquisition (Driver, Newton, & Osborne, 2000; Schwarz, Neuman, Gil & Ilya, 2003; Zohar & Nemet, 2002), and epistemological understanding of knowledge construction (Duschl & Osborne, 2002; Erduran, Osborne, Simon, 2004; Sandoval & Reiser, 2004). However, several researchers have pointed at the difficulty teachers have at mediating collective argumentation. They need to continuously evaluate students' knowledge and to enact argumentative moves timely and adequately in order to promote understanding: asking for new perspectives, pointing at contradictions, inviting to participate, inviting to give explanations, etc. (Yackel, 2002; Schwarz, Dreyfus, Hershkowitz, & Hadas, 2004). The complexity in sustaining argumentation has led scientists to elaborate CSCL tools for supporting collective argumentation in peer interaction. These tools (also called CSCA tools) have to be differentiated from knowledge representation tools (Bell, 1997; Van Bruggen, Kirschner, Jochems, 2002), since they support collective argumentation by affording argumentative moves to be taken by various speakers through discussion. The Digalo environment (Glassner & Schwarz, 2005) presents an approach that integrates these two models of fostering argumentation: the model emphasizing "knowledge representation", or "argument representation", and the supportive model in the argumentative process. Using this tool synchronously enables textual multiple-talk through which each of the subjects adds messages through mediation of graphical icons representing categories in collective argumentation and in argument construction (these categories are called the ontology of the environment). The underlying assumptions of the designers of this tool were that visual ongoing representation of the discussion can help students to reflect upon their argumentative steps and their components, and that discussants will enact practices of productive discussion (Glassner & Schwarz, 2005). And indeed, synchronized communication amongst subjects, mediated by ontologies characterizing collective argumentation (such as 'claim', 'argument', 'explanation', 'comment', 'question') as well as relational categories (such as 'support', 'opposition' or 'reference'), appeared to be productive (Glassner & Schwarz, 2005). In other words, adequate ontologies may have a mediating effect in the sense that when learners use the tools, they take more into consideration crucial features of collective argumentation, among them the reference to others as well as criticism and justification. The persistence of the argumentative map suggests the clear articulation of opinions, inspection of the map to decide whether the contribution is new, reflection to evaluate understanding of previous moves, and to possibly request additional explanations, etc. However, these suggestions have been empirically corroborated without considering the role of human moderation, a role that was recognized as crucial in the long run (Shahar, 2003; Hakkarainen, Lipponen, & Järvelä, in press). It is then imperious to focus on the role of human moderators in helping students while discussing with CSCL tools. To begin with in this endeavor, we use the general term 'moderation' to designate any kind of support given by a human to help at reaching the goal of the e-discussion. Researchers used different terms such as 'scaffolding', or 'mediating' to delineate teaching actions aimed at supporting construction of knowledge. Since we adopt a bottom-up approach, we will use 'moderation' as a general term but we will preserve terms used by researchers Three types of mediations have been identified in CSCL literature for either synchronous or a-synchronous interactions (Ashton, Roberts, & Teles, 1999): pedagogical scaffolding; social scaffolding; and technological scaffolding. Pedagogical scaffolding refers to moves aimed at achieving predetermined learning goals in order to help the learner complete his/her assigned task (Mercer, 1995; Muukkonen, Hakkarainen & Lakkala, 1999; Wegerif, 1996). It includes positive reactions, instructions, providing information and opinions, advising, pointing out preferences, raising questions, and a summary of students' remarks concerning external sources. Various researchers adopted different strategies. For example, Chi and colleagues (Chi et al., 2001) proposed an approach according to which interaction between teacher and student can be planned by asking generic questions, such as: “Can you explain...”, or “Articulate it with your own words”, “What are you thinking about the issue?”, “Could you add anything about the subject?” (See also Baker and Lund, 1997, for a similar approach that yielded positive outcomes). While Wegerif (1996) proposes a similar approach with scaffoldings in the form of questions such as: "What are you thinking?", and "Why are you thinking that way?", his perspective is less cognitive than ethical and dialogical since he is eager to instill ethical norms of argumentation rather than to instigate dialectical processes. Social scaffolding refers to support based on empathy, humor, and personal assistance. Technical scaffolding includes technical support with the software and interface for students working in a CSCL environment. These and other kinds of distinctions concerning kinds of moderation were done either theoretically or by observing teaching practices. In this paper, we adopt a bottom-up approach as we both consider both beliefs and practices of teachers in a specific case – synchronous e-discussions, so as to distinguish between kinds of moderation in that case. The nature of human moderation in synchronous e-discussions and its influence on the quality of such discussions are open questions. These questions are relevant to ARGUNAUT (IST-2005027728 – partially funded by the EC under the 6th Framework Program, http://www.argunaut.org), a project aimed at providing tools for supporting the moderation of synchronous e-discussions. The present paper represents an initial step in this program. First of all, we need to learn about teacher beliefs and practices concerning moderation of synchronous discussions. The first study contrasts teachers' beliefs concerning good discussions and good moderation in face-to-face discussions and in synchronous, CSCL-mediated discussions. The second study focuses on the actual strategies teachers intuitively enact in synchronous discussions. In the two studies, the populations were different. This was due to difficulties in recruiting and training teachers to be moderators in synchronous discussions. The results of the two studies suggest the elaboration of suitable awareness tools to help teachers in mediating CSCL discussions. The Research questions 1. How do teachers characterize and describe what constitutes a good discussion in oral classroom settings, as opposed to discussions in CSCL-based argumentation? 2. How do teachers characterize and describe what constitutes good moderation of oral classroom discussions, as opposed to moderation of discussions in CSCL-based argumentation? 3. How do teachers actually moderate synchronous, collective argumentation? Study 1: Teachers beliefs about good discussions and moderation of discussions