Effects of peer-challenge support on learning during on-line small group discussion

by Choi, Ikseon.

Meaningful discussion helps learners to construct their own knowledge by providing several cognitive benefits such as articulation, cognitive conflicts, and co-construction (Crook, 1994; Jonassen, Peck, & Wilson, 1999; Koschmann et al., 1996; Tao & Gunstone, 1999b). Although on-line discussion has been used with these expectations for learning benefits, the actual effects are unclear. Meaningful discussion can be initiated when learners raise thoughtful questions or provide critical feedback; however, generating effective questions requires a certain level of domain knowledge and metacognitive skills of the questionaskers (Dillon, 1986; Miyake & Norman, 1979; Palincsar & Brown, 1984; van der Meij, 1990; Wong, 1985). Unfortunately, novice learners who begin to explore a new area are often limited in those metacognitive skills (Land, 2000; Garner & Alexander, 1989), so they may experience difficulties in generating meaningful interactions. In cases where simple question-answer cycles are employed in a discussion, learners may not benefit from the interactions in a way that enhances effective knowledge construction. This study proposes a peer-challenge support framework intended to facilitate effective peer interactions in on-line discussion and thus enhance learning. This framework assumes that novice students who lack domain knowledge and metacognitive skills can be supported in generating meaningful interactions at an early stage of learning (King & Rosenshine, 1993; Palincsar & iv Brown, 1984). In turn, the resulting questions and feedback can enhance peers' metacognition, such as reflecting and monitoring, which can allows them to refine and restructure their understanding (Piaget, 1985; Webb & Palincsar, 1996). Thus the purpose of this study was to test this peer-challenge support framework by investigating the effects of providing externalized, on-line support for generating effective peer-challenges during on-line small group discussion, thereby enhancing learning in college students. On-line peer-challenge support included descriptions of what to do and generic and domain-specific examples of questions. A field experimental time-series control-group design was employed as a mixed model for the research design. Thirty-nine students from an on-line introductory class on turfgrass management offered by a large northeastern landgrant university participated in the study. The participants were randomly assigned to a small group of four to six members. Then those small groups were randomly assigned to either an experimental or a control group. Quantitative and qualitative data were collected from five successive sessions of on-line small group discussion, ten sessions of open-ended essay exams, three sessions of multiple-choice exams, an on-line survey, and a telephone interview (of four volunteers) throughout a sixteen-week semester. The quantitative results revealed that on-line peer-challenge guidance helped learners to generate significantly more challenges. However, the guidance did not improve the quality of challenges, further interactions, and learning v outcomes. The increased quantity of challenges alone might not be sufficient to improve meaningful interactions. Based on the survey data, possible reasons for the failure of showing the guidance effects on learning might be the following: students’ limited use of the guidance, their relatively high level of prior experience, and a small sample size. The qualitative results based on the interview data were consistent with the peer-challenge support framework. The results verified that students tended to experience difficulties in generating challenges for their peers when they perceived a lack of knowledge. Importantly, they perceived that the use of online guidance helped them to improve their questions in both quantity and quality when they had difficulties in generating challenges. Furthermore, the students who received meaningful challenges from their peers tended to experience cognitive dissonance which triggered them to reflect upon their understanding, to articulate their lack of knowledge, and to seek necessary information until they built up enough knowledge to be able to generate satisfying answers. vi