Pattern Discovery for the Design of Face-to-Face Computer-Supported Collaborative Learning Activities

Article excerpt

Introduction

According to constructivist learning theory, collaboration and interaction between peers promote deep learning by exposing students to different perspectives and providing opportunities for negotiation (Brett & Nagra, 2005). This effect is reinforced when the students' activities are supported by computers (Africano et al., 2004). Computersupported collaborative learning (CSCL) is a research field that studies how to mediate interactions within groups in a technological environment (Roschelle & Teasley, 1995). CSCL methods provide a number of mediation elements that support learning and social interaction among members of a collaborative group working on a particular learning task (Zurita & Nussbaum, 2004). Given that social interaction is fundamental to effective learning (Johnson & Johnson, 1987; Stanton et al., 2001; Vygotsky, 1974), various technology-supported learning environments have been created to mediate such interaction. When the computers employed in a classroom are wirelessly interconnected with personal digital assistants (PDAs), the result is a collaborative one-to-one computing work environment (Liu & Kao, 2005) that enables face-to-face CSCL interaction between peers (Zurita & Nussbaum, 2004), thus facilitating coordination, communication, and interactivity among group members (Shin, Norris, & Soloway, 2006). In face-to-face CSCL environments, collaboration and communication are conducted across wireless and social networks. Thus, while users are communicating face to face via the social network, they are simultaneously supporting their work through the wirelessly interconnected handheld network, the result being an easy exchange of information between peers or between student(s) and teacher (Pea & Maldonado, 2006).

Since free collaboration does not necessarily produce learning (Dillenbourg, 2002), it is essential that the collaborative learning process be performed in an appropriate manner. One way of improving the interaction/collaboration process is to structure interactions by engaging students in well-defined scripts. A collaboration script is a set of instructions prescribing how students form groups, interact and collaborate, and solve the assigned problem (Kobbe et al., 2007). This is consistent with Engestrom's activity theory (1987), which posits the need for rules and regulations for the mediated relationship between subject and object. These rules and regulations define individual and group responsibilities, coordination and organization mechanisms, and roles to help ensure the activity aim is achieved (Adams & Hamm, 1996; Dillenbourg, 1999).

If the basic unit for modeling actors within a classroom is the interaction among them (Carroll, Neale, Isenhour, Rosson, & McCrickard, 2003), the study of interaction will then be the key to understanding the process of individual and group learning (Lagos, Alarcon, Nussbaum, & Capponi, 2007). The aim of the present work is to determine, for collaborative learning inside the classroom, which social interactions foster and which ones hinder the achievement of the defined learning activity objective.

In the methodology proposed here, we cluster the social interactions that occur within the classroom among peers and between students and teacher into patterns. According to Alexander, Ishikawa, and Silverstein (1977), patterns constitute a format for sharing practices so that they can be used by a variety of people in many different ways (Eckstein, Lynn, & Voelter, 2001). Education technologists have looked for patterns to solve problems in collaborative activities in CSCL systems (Hernandez-Leo, Asensio-Perez, & Dimitriadis, 2004; Baggetun, Rusman, & Poggi, 2004) and to capture teaching practices in their respective disciplines (Eckstein et al., 2001; Bergin, 2002; Frizell & Hubscher, 2002). For example, the Pedagogical Patterns Project (Eckstein et al. …