Theory-Driven Natural Language Processing: Revealing Distinct Patterns of Pedagogical Communication in Algebra Tutoring

• In Event: Technology, STEM Learning, and Contexts

Abstract

This study investigates five tutors’ communication patterns in an online algebra learning environment (Algebra Nation) to provide a framework for understanding pedagogical communication strategies in video lectures. The corpus comprised 125 algebra tutoring videos from Algebra Nation, which were analyzed using natural language processing. Three functional categories of teacher-talk — Interpersonal, Structural, and Cognitive — were used to guide the analyses and interpret the findings, providing a customized evaluation and analytic framework of the tutors’ communication strategies in the video lectures. The findings revealed that tutors adopt several distinctive communication patterns to effectively manage their video lectures. Theoretical interpretations of the findings are discussed based on pedagogical discourse patterns’ functional paradigms.

Authors

• Jinnie Shin, University of Florida

• Renu Balyan, Arizona State University

• Michelle Banawan, Arizona State University

• Walter L. Leite, University of Florida

• Danielle S. McNamara, Arizona State University