Identifying Topics and Trends in U.S. Policy/Vision Documents Using Latent Dirichlet Allocation (Poster 21)

• In Event: Thursday Poster Session 12:40 pm
  In Poster Session: Mathematical Teaching and Learning Poster Session

Abstract

In this study we used latent Dirichlet allocation (LDA) to investigate topics in a collection of U.S. policy/vision documents beginning with the Committee of Ten, moving to the New Math era, continuing with the back-to-basics era, and ending at the standards era. We compared the topics we identified in these documents using LDA to summaries of these documents produced by the Center for the Study of Mathematics Curriculum. Our analysis led to the identification of new topics that had hitherto not appeared in these documents or in summaries of these documents such as linking problem solving to rigorous arithmetic or parents as tools for students to learn mathematics. When looking across document topics we found four different themes.

Authors

• Jon D. Davis, Western Michigan University

• Paul H. Atandoh, Mercer University