Exploring the affordances of Bayesian networks for modeling usable knowledge and knowledge use in teaching

Nicole B. Kersting, James E. Smith, Beau Vezino, Mei Kuang Chen, Marcy B. Wood, James W. Stigler

Research output: Contribution to journalArticle


In this article we propose the use of Bayesian networks as a potentially promising way to model usable knowledge. Using the Classroom Video Analysis (CVA and CVA-M) assessments as a lab model for studying teachers’ usable knowledge, we first explored whether we can identify the knowledge (pieces) underlying teachers’ written responses. In the CVA approach we ask teachers to respond to short video clips of authentic classroom instruction based on different prompts that are designed to simulate common teaching tasks. We then explored the affordances of Bayesian networks to functionally model usable knowledge as an interconnected dynamic knowledge system consisting of different knowledge pieces and connected pathways weighted by situation-specific relevance and applicability. We explore the implications of these models for studying the development and growth of usable knowledge and propose the use of Bayesian networks as a novel and potentially promising way to model usable knowledge and for understanding how knowledge is used in teaching.

Original languageEnglish (US)
Pages (from-to)207-218
Number of pages12
JournalZDM - Mathematics Education
Issue number2
StatePublished - May 1 2020



  • Bayesian networks
  • Knowledge system
  • Knowledge use
  • Teacher knowledge
  • Teaching practice
  • Usable knowledge

ASJC Scopus subject areas

  • Education
  • Mathematics(all)

Cite this