How Teachers Talk Algebra Online: Linking Discourse Patterns, Instructional Experiences, and Student Outcomes
Digital learning platforms have transformed mathematics teaching and learning, placing unprecedented importance on the quality of teacher discourse in video lectures, a medium of instruction devoid of physical interaction commonly found in traditional classrooms. Using the systemic functional linguistics framework, we conducted a fine-grained analysis of 100 video lectures on 25 algebra topics produced by four teachers with varying backgrounds, examining how they represent concepts and ideas, structure explanations, and foster engagement. Our findings reveal that patterns of teacher talk vary significantly, influenced more by experience in creating online lessons than by years in the traditional classroom. Specific features of teacher talk were strongly associated with student outcomes, measured by in-video dropout rate and accuracy rate on post-lesson assessments: Discourses that attempted to clarify concepts, elaborate ideas, compact information, engage students, and build community positively impacted student outcomes, whereas discourses that were overly technical, additive, or lengthy negatively affected student outcomes. These findings suggest that teacher discourse can shape students’ ability to understand content, apply knowledge to problem solving, and persist in their learning. This research provides valuable insights into the linguistic architecture of algebra instruction and can inform efforts to improve online learning through the production of quality video lectures.
Authors
- Zifeng Liu (ORCID: https://orcid.org/0009-0005-5833-2141)
- Wanli Xing (ORCID: https://orcid.org/0000-0002-1446-889X)
- Zhihui Fang (ORCID: https://orcid.org/0000-0002-1822-4331)
- R. S. Guo (ORCID: https://orcid.org/0000-0002-1140-6765)
Publication Details
- Journal
- Journal of Learning Analytics
- Published
- 2026-11-15
- DOI
- https://doi.org/10.18608/jla.2026.9455
- Primary Topic
- Mathematics Education and Teaching Techniques
- Type
- article
- Field-Weighted Citation Impact
- 0.00