Geographic Patterns of Math Achievement: A Methodological Case Study of Regression Modeling Approaches (Poster 31)

• In Event: AERA Poster Session Thursday 8:00 a.m.
  In Poster Session: Advancements in Educational Data Analysis: Machine Learning, Multilevel Models, and Beyond

Abstract

This study investigates how different regression models with spatial and clustering components account for variation in mathematics achievement based on the characteristics of students, teachers, and regions. The field of educational research often ignores geographic diversity, potentially resulting in models that overestimate the importance of other contexts. By incorporating spatial information into a cross-classified multilevel model, this study examines whether inferences about the student, teacher, and regional factors that affect math achievement can be improved. The results of this study could provide valuable insights into the role of geographic variables in educational outcomes.

Authors

• Lodi Lipien, Florida Virtual School

• John M. Ferron, University of South Florida

• Eunsook Kim, University of South Florida