Comparison of Predictors Relating to Teacher Burnout by School Level via Post-Selection InferenceComparison of Predictors Relating to Teacher Burnout by School Level via Post-Selection Inference

The purpose of this study was to explore large-scale educational data with penalized regression, and to investigate significant predictors to teacher burnout by school level. Compared to previous studies, this study has the advantage of encompassing and analyzing teacher-related variables at the individual, organizational, and societal levels across elementary, middle, and high schools. Specifically, the Seoul Teachers Longitudinal Study (2020) 1st wave data was analyzed via Post-Selection Inference (PSI). After 100 iterations of PSI modeling, significant predictors were compared across school levels. As a result, predictors including teaching experience and teacher’s fixed mindset on students’ intelligence development potential were commonly selected across all school levels. Meanwhile, predictors of parental counseling duties and the economic level of the school location were selected only in elementary schools and middle schools, respectively. Policy recommendations to effectively address teacher burnout were discussed based on the findings.