Factors Affecting Teachers’ Intention to Leave: An Exploratory Study using Machine Learning and Logistic Regression

• In Event: AERA Poster Session Sunday 1:45 pm
  In Poster Session: Methodological Techniques for Challenging Data and Modeling Situations

Abstract

This study explores factors influencing public school teachers’ intention to leave their jobs using nationally representative data from the 2023 RAND State of the American Teacher Survey. We employed an integrated approach combining machine learning and logistic regression with a range of predictors, such as working conditions, administrative support, workplace autonomy, and teacher and school characteristics. The machine learning model identified key variables predicting teacher intent to leave the profession, while the logistic regression confirmed the significant effects of these predictors, particularly satisfaction with working hours, school safety concerns, and administrative support. By integrating data-driven and inferential methods with the up-to-date data, this study offers both methodological and practical insights for addressing teacher turnover in U.S. public schools.

Authors

• Tzu-Wei Wang, Virginia Commonwealth University

• Michael Broda, Virginia Commonwealth University

• Georgia Heyward, Virginia Commonwealth University

• Cammie Justus-Smith, Virginia Commonwealth University