Data Tools for Educators: Studying Practitioners' Reactions to Data-Driven Student Groupings Suggested by Machine Learning

• In Event: Deepening Our Understanding of Teacher Data Use: Antecedents, Forms, and Supports

Abstract

Despite unprecedented levels of datafication in education, little is known about the tools practitioners need integrate data into the day-to-day tasks of working with students. One way to close this gap is to focus on ways to analyze and present data for specific pedagogical tasks such as group formation. Accordingly, this study investigated how educators (n = 20) reacted to and used data-driven student groupings suggested by a machine-learning algorithm. Results suggest two factors influence their willingness to “lean on” data-driven groupings generated by machine learning: the type and complexity of the underlying student data. The paper discusses the implications for researchers and practitioners interested in empowering educators to use data more effectively, while acknowledging the limitations of the study.

Authors

• Daniel L. Hoffman, University of Hawaii - Manoa

• Mark Santolucito, World Scholars LLC

• Maria Lara Hwang, Teachers College, Columbia University

• Seungoh Paek, University of Hawaii - Manoa