Exploration and Identification of Predictors to Teacher Job Satisfaction via a Machine Learning Technique, Group Mnet

• In Event: Hilton Roundtable Session 6
  In Roundtable Session: Machine Learning and Data Sciences

Abstract

The current study explored and identified important predictors to teacher job satisfaction with TALIS data. Specifically, group Mnet was employed as a machine learning technique to explore the hundreds of variables in one prediction model. Mnet executes variable selection with consistency and handles multicollinearity. Additionally, group Mnet treats dummy-coded variables from the same categorical variable as a group in variable selection. A total of 548 variables from the teacher and principal TALIS questionnaires were explored after data cleaning and merging, and 18 variables were identified as important after relevance counts. Among the 18 variables, 7 of them had been studied in the previous research, and the other 11 were newly identified variables. The scientific importance of this study was discussed.

Authors

• Jin Eun Yoo, Korea National University of Education

• Minjeong Rho, Korea National University of Education