Toward the automated analysis of teacher talk in secondary ELA classrooms

Qualities of teacher talk strongly affect student learning, yet opportunities for teachers to receive feedback on their talk face time and cost limitations. To address this, we developed a semi-automated method of analyzing teacher talk using Automated Speech Recognition technology, compared it to traditional methods, and used it to analyze teacher talk features in 127 secondary English Language Arts lessons. We found the new method aligned with traditional methods, allowing us to reliably identify interrelated but distinct talk features. These features often reflected dialogic instruction. Our findings contribute to future studies of instructional discourse and technology-driven teacher learning opportunities. (c) 2021 Elsevier Ltd. All rights reserved.

Automated summary

The quality of teacher talk significantly impacts student learning, but opportunities for teacher feedback are limited. A semi-automated method using Automated Speech Recognition technology aligned with traditional methods, reliably identifying teacher talk features reflecting dialogic instruction. These findings contribute to future studies of instructional discourse and technology-driven teacher learning opportunities.