Investigation on Students’ Sense of School Belonging: Application of Mixed-effects Random Forest to TIMSS Data

• In Event: Educational Research in Transition: From Classical Methods to Machine Learning

Abstract

There has been insufficient research on students’ sense of school belonging for elementary students despite of its importance. This study aims to explore predictors on early adolescence students’ sense of school belonging via machine learning techniques. The authors employed mixed-effects random forest to handle hierarchy of data. Of the 172 TIMSS variables explored, 20 student variables were presented as important for predicting elementary school students’ sense of school belonging. The study results can shed light on predicting students who have low sense of school belonging and making timely interventions for them. Further research topics were also discussed.

Authors

• Miryeong Koo, University of Illinois at Urbana-Champaign

• Jinming Zhang, University of Illinois at Urbana-Champaign