Using Bayesian Networks to Model Teachers' Usable Knowledge and Knowledge Use in Teaching

• In Event: Marriott Roundtable Session 2
  In Roundtable Session: Modeling and Developing Usable Knowledge of Mathematics Teaching

Abstract

In this study we explore the usefulness of Bayesian networks as a novel approach to functionally describe teachers’ usable knowledge and knowledge use in mathematics. We use teachers’ responses to the Classroom Video Analysis (CVA-M) measure as a proxy measure of usable knowledge. We identify the different knowledge pieces contained in teachers’ CVA-M responses and use this information as input to discover the structure and associated weights of the usable knowledge network. Results indicate that early carrier and experienced teachers access similar knowledge in their responses, but that, as expected, the structure of their knowledge is qualitatively different, providing promising preliminary empirical evidence for the usefulness of computational models to describe and study useable knowledge and knowledge use in teaching.

Authors

• Nicole B. Kersting, The University of Arizona

• Beau Vezino, Northern Arizona University

• James E. Smith, University of Arizona

• Mei-Kuang Chen, The University of Arizona

• Marcy B. Wood, The University of Arizona

• James W. Stigler, University of California - Los Angeles