Predictors of Teacher Retention in Chile Light-Gradient Boosting Model with Annual National Data (2017 cohort).pdf

• In Event: AERA Roundtable Session Wednesday 4:20 pm Four Seasons Ballroom 1
  In Roundtable Session: Teacher Retention Across Contexts (Table 13)

Abstract

This study investigates the predictors of teacher retention in Chile using a Light-Gradient Boosting Model with national data from 2017. A cohort of 10,926 teachers was constructed and tracked over seven years. Key factors influencing teacher retention included contract type, contract hours, and the level taught. Significant predictors of the number of years teachers remain in the system were having a permanent contract, more contract hours, and teaching at the elementary level. The study revealed a 29.4% attrition rate by 2023, exceeding previous estimates. These findings highlight the critical role of job security and working conditions in enhancing teacher retention and offer valuable insights for policy-making to reduce turnover and improve educational stability.

Author

• Cristobal Manaut, Universidad de Chile