The Relationship Between Academically Productive Talk and Instructional Quality in Mathematics Lessons

• In Event: Instructional Practices in Mathematics Classrooms

Abstract

This paper investigates the relationship between teacher and student discourse patterns, measured by accountable talk moves (Michaels & O’Connor, 2015) and the quality of mathematics instruction as measured by the Mathematical Quality of Instruction (MQI) rubric. This study uses a large public dataset of human coded MQI lesson transcripts and validated AI coding for talk moves to explore how different talk moves predict instructional quality. Results indicate that certain talk moves at certain frequencies, especially those relating to accountability to the learning community and rigorous thinking, positively correlate with higher MQI scores. Thus the nature and frequency of nuanced discourse patterns are crucial for high-quality mathematics instruction, while simple metrics like the amount of student talk have little impact.

Authors

• Jennifer K. Jacobs, University of Colorado - Boulder

• Kaylee Thomas, University of Colorado - Boulder

• Mimi Engel, University of Colorado - Boulder

• Jeffrey Bush, University of Colorado - Boulder