Developing students’ understanding of productive classroom talk using Classroom Discourse Analyzer

• In Event: Challenges and Possibilities of Teaching Innovation and Curriculum Reform

Proposal

Background: Globalization and multiculturalism ask for new approaches to communication and problem-solving. In a connected but diverse world, negotiation and discussion are employed not just to summarize what leaners already learn, but also to improve ideas and craft promising solutions to real-world and ill-defined problems. An emerging body of work in linguistics, psychology, education technology discusses how to engage learners in productive inquiry from social-cognitive and technological perspectives. “Deliberative democracy”, a notion emphasizes both individual freedom and group solidarity has been taken up by a range of theorists. Deliberative democracy has long been linked to promotion of dialogue and discussion in the field of education (Deway, 1966; Habermas, 1990, Resnick et al., 1993). A large number of studies try to improve classroom discussion through categorizing various types of dialogue and identifying “good moves” in classroom talk (Bereiter & Scardamalia, 2016). Several representative analytic frameworks, such as “accountable talk”, “knowledge-creating dialogue” and “educational dialogue” have been developed to promote students’ use of productive talk moves (Alexander, 2018; Hennessy et al., 2016; Lei, 2018; Resnick et al., 2010). Researchers point out productive talk moves build up understanding through socially shared questions, answers, and feedbacks. Multiple studies uncover that reflection on the quality of discourse is an effective strategy to foster productive classroom talk. To help learners review previous collective inquiry process, researchers design visual learning analytic tools through evidence-based classroom discourse analysis. The Classroom Discourse Analyzer (CDA) is a tool visualizes and synchronizes basic classroom talk information (e.g., transcripts) and high-inference information (e.g., category of discourse moves) (Chen, et al., 2020). Two research questions were addressed in this study: (1) How students reflected on the visualized classroom discussion data supported by CDA? (2) How reflection on the visualized CDA data influenced students’ use of productive talk moves?
Methods: Participants included one class first year undergraduate students (n=25) from a university in Shanghai, China. Students were divided into five groups, and each group had 5 students. They attended weekly class and participated in group discussion over a period of 12 weeks. Each week, one open-ended authentic question was provided after instruction of domain knowledge, and students were given 10 minutes to thoroughly discuss the topic with group members. The teacher had more than twenty years of teaching experience and English was used as the medium of instruction and discussion. The course conducted during this study, entitled Introduction to Business, was designed to help first-year undergraduate students understand basic business concepts and theories. At the first week of the semester, students participated in a pre-domain test which examined their conceptual understanding of business. And the result showed there was no significant difference in students’ domain knowledge before class. We used a randomized controlled trial with an embedded case study to examine how reflection enriched with CDA affected students’ classroom talk. Three focus groups of students (n=15) were invited to participate in reflective discussion based on the CDA; while the other two groups of students (n=10) were the comparison groups who joined the regular classroom discussion without further reflection. Classroom discussion and collective reflection on CDA data were recorded and transcribed afterwards. The intervention students first learned about coding scheme about “productive classroom talk” in an introductory workshop, then participated in three one-hour video-based workshops (once every four weeks) across the semester to reflect on their classroom discussion. The coding scheme included two dimensions: (1) regular talk moves that foster individual inquiry (e.g., share own idea), (2) productive talk moves that foster collective inquiry (e.g., synthesize group idea). Students navigated visualized data by clicking bubbles, diagrams, video clips, and transcripts. They collectively reviewed their use of talk moves to reflect how they engaged in classroom discussion and identify the episodes of interest. Several scaffolding questions were designed to promotes students’ reflection, such as “What happened in this clip?”, “What impressed you when you browse the diagram on the CDA?”, and “What changes do you expect to have in your classroom talk in the next month?”. Discourse analysis was used to investigate how students reflected on the CDA visualizations and how they used talk moves in follow-up classroom discussion.
Findings: In the first investigation, students’ reflection on CDA visualizations could be divided into three levels. Low-level reflection was able to identify their classroom discourse pattern, but some ideas were inconsistent with the CDA visualizations or it missed some information presented by the CDA. Middle-level reflection was more advanced in accuracy of discourse pattern identification. Also, students in middle level could further explore information on individual learner’s performance by clicking the bubbles and contextualizing code in origin transcriptions. High-level students precisely described discourse patterns and have more sophisticated understandings of the whole picture of group discussion. For example, high-level reflection explained how group collectively connected and improved ideas during journey of inquiry, identified the gaps of the use of talk moves, proposed future direction of improvement. Three groups continuously improved the quality of reflection in the last two workshops. The second question examined whether intervention groups were able to employ more productive talk moves in follow-up group discussion. Discourse analysis uncovered that intervention groups increased use of productive moves after reflection enriched with CDA. Three types of productive talk moves were only identified in intervention groups: (1) meta-collective reflection of state of knowledge, (2) meta-collective advancing of idea improvement, (3) meta-collective deepening and emergent problems.
Contribution: Two major affordances of reflection on CDA data may contribute to students’ awareness of productive talk moves. First, evidence-based and process-oriented visualizations help students navigate and select data for reflection. Second, CDA tool helps students recognize how they collectively and continuously improve their group inquiry and how their group discussion resembles or differs from other groups, which empowers meta-cognitive discussion and reflection.

Author

• Junjie Zhang, The University of Hong Kong