Epistemic Development in Small Groups

The analysis of the participants’ interactions shows that learners followed certain participatory pattern when co-constructing epistemic knowledge in their small groups.

As indicated by previous research in productive interaction, various aspects are documented to influence the productivity and epistemic development in group collaboration. More specifically, the interplay among participants’ active participation, sharing and construction of knowledge, emergent knowledge objects, and the coordination and monitoring of group collaboration plays an important role in creating productive interactions and better learning outcomes (Barron, 2003).

In the present study, examples demonstrating epistemic development at group levels include episodes of group members achieving group understanding, co-constructing knowledge, and resolving conflicting perspectives. As demonstrated in Figure 2, the participatory pattern in these episodes starts with actions in category 1 &

2 (the emergence of problems, the identification of focus, or the sharing of knowledge).

Subsequent to the establishment of common ground, actions in category 3 & 4, including those aiming at generating, elaborating, and negotiating ideas, follow for the purpose of having productive interactions and accomplishing the assigned tasks. In other words, a trajectory of epistemic development, which starts from knowledge sharing, the emergence of problems, the negotiation of solutions, to the development and construction of new knowledge is observed.

69

Figure 2 Recursive pattern in group collaboration

The identified pattern reflects different interactional phases in group learning, which is similar to what Onrubia and Engel (2009) have observed in their study with students engaging in collaborative writing task through virtual classrooms. Four phases of knowledge engagement are identified in their study, including phase of initiation, exploration, negotiation, and co-construction. In accordance with their findings, knowledge is accumulated and refined as group discussions unfold. In addition, a gradual involvement with knowledge and continuous knowledge co-construction are found, which resonates with the idea that the creation of shared understanding in small group requires ongoing elaborations and should be captured from a more dynamic view across temporal and sequential unfolding of tasks (Damşa, 2014; Teasley, 2008).

It should be noted that the identified interactional pattern is not a linear, but a circular one. That is to say, subsequent to the actions in category 3 & 4, actions in category 1 & 2 might emerge if necessary and beneficial for the group to complete the tasks. As our participant, Alex, said in the interview, “When encountering problems that we were not able to solve even after group discussion, we need to search for more information for the purpose of coming up with a solution that every group member is satisfied with.”

The findings in our study also show that convergence of ideas often occurrs after group members have shared knowledge and negotiated points of view and that

70

agreements might come late in the whole discussion (Gunawardena et al., 1997;

Kimmerle et al., 2017). Our results coincide with previous findings and reinforce the idea that the development, reconciliation, and consolidation of knowledge could be successfully achieved through collaboration and peer support.

In terms of the construction of language-related knowledge, the most common interactional pattern observed in our data is composed of short back-and-forth interactions from the actions of category one (stating problems) to those of category two (sharing knowledge). Such interactions constituted by one-directional sharing are more salient when group members at higher English proficiency level play a dominant role, which resonates with the findings that proficiency level is a crucial factor in influencing the interactional patterns in the context of dyad learning (Watanabe and Swain, 2007) as well as in small group learning (Wen et al., 2015). In other words, sharing or elaboration, instead of negotiation, of word use is more likely to be found in the interactional moves within a mixed-ability group where students at higher proficiency level dominate the conversation.

Still, compared with group interactions in face-to-face settings, those facilitated by the use of technological tools show a more balanced and increased participation for each group member, as suggested by previous studies (Vetter and Chanier, 2006). In traditional language classrooms, students often draw on their linguistic knowledge to undertake tasks at hand, which unavoidably leads learners at higher proficiency levels to guide or even dominate the conversation. The situation and the unbalanced participation among group members are more likely to be mitigated in technology-supported and Internet-connected language classrooms where language learners at varying proficiency levels are armed with opportunities and access to explore linguistic knowledge and language use whenever they want. As shown by the action of one participant checking the accuracy of his own answers (Excerpt 2), the participants are

71

able to use mobile devices at hand to cull accurate information regarding language use.

In this way, mobile devices support language learning for both individuals and groups by offering immediate and personalized support. As suggested by our participants in the interviews, the use of iPads makes it easier for them to look up unfamiliar words and thus promotes engagement in group discussion.

From the interaction analysis, peer scaffolding is proved to be crucial in the collaborative learning process. Our excerpts demonstrate how students develop their knowledge with the help of learning peers. Vygotsky, in his social constructive theory of knowledge, has acknowledged the role of “more knowledgeable other” in fulfilling learning potentials, the so-called zone of proximate development (ZPD), of less-competent learners. The idea of more capable learners helping and promoting growth is supported by the findings of Kirsch (2006) that more knowledgeable children facilitate their peers’ learning by encouraging talk and promoting comprehensions in the tasks of digital storytelling. Based on the analysis of Excerpt 3, however, it is interesting to find that the “more knowledgeable others” do not always refer to the same person or group of people; each group member involving in the discussion can be a potential source of knowledge. In other words, peer scaffolding emerges dynamically and bi-directionally. In Excerpt 3, the ostensible less knowledgeable person who raises a vocabulary question in the beginning is able to build on the contribution of others and becomes a source of knowledge, which facilitates small group learning. It seems that through social interactions, ideas are communicated verbally and are thus more likely to be fine-tuned or transformed into a rather internalized piece of knowledge (Vygotsky, 1986).

The benefits of peer scaffolding and collaboration are also confirmed in the participants’ learning journals and interviews. For example, each group member might bring in their own background knowledge or pay attention to different parts while

72

constructing knowledge collaboratively, through which all of the group members can gain insights into the topics being discussed (Journal, Claire, 1030; Journal, Elva, 1002).

Previous studies have shown that conflicts, or conflicting perspectives, are helpful in triggering deeper exploration of the topic (Jermann & Dillenbourg, 2003). The process of resolving conflicting ideas is also considered to be one of the driving forces for cognitive change (Mugny, Perret-Clermont, & Doise, 1981). In short, collaboration and the emergence of conflicting perspectives in small group learning creates diverse opportunities for the participants to construct knowledge in a more comprehensive manner.

73