A visual learning analytics (VLA) approach to video-based teacher professional development: Impact on teachers’ beliefs, self-efficacy, and classroom talk practice

To address the challenge of overwhelming data inherent in classroom lesson videos, this study proposed a visual learning analytics (VLA) approach to video-based teacher professional development (TPD). Using a two-year experimental design, 46 secondary mathematics teachers were divided randomly into a treatment group (N = 24) and a control group (N = 22) to learn about and integrate academically productive talk into their teaching. The treatment teachers participated in a VLA-supported TPD program, while the control teachers participated in conventional knowledge-based workshops. Results show that teachers in the treatment group had more positive beliefs and higher self-efficacy in the post-test and delayed-post-test, while the control group improved, but not significantly, in their beliefs about the usefulness of classroom talk. In addition, although the control group made a significant improvement in their self-efficacy in guiding classroom talk in the post-test, this improvement was not sustained to the delayed post-test. Moreover, the coding of classroom teaching behaviour revealed that teachers in the treatment group relative to the control group significantly increased their use of academically productive talk in the post-test lessons to encourage the students' elaboration, reasoning, and thinking with others in the classroom. The results suggest that, while attending knowledge-based workshops had, to some degree, positive effects on the control teachers' beliefs and self-efficacy, these effects were not sustainable over time. In contrast, the use of visual learning analytics to support the treatment group's reflection on the classroom data not only had significant and sustained effects on the teachers' beliefs and self-efficacy but also significantly influenced their actual classroom teaching behaviour. Implications for designing VLA to support teacher learning and professional development are discussed. •A VLA approach to TPD was found to be effective in supporting teacher reflections.•Control teachers' beliefs and self-efficacy about classroom talk improved.•Control teachers' improvement was not sustained to the delayed post-test.•Treatment teachers' beliefs and self-efficacy improved significantly and sustained.•Treatment teachers significantly improved classroom talk strategies in teaching.