A machine learning approach to identifying key predictors of Peruvian school principals' job satisfaction DOI Creative Commons
Luis Alberto Holgado-Apaza, Dany Dorian Isuiza-Pérez, Nelly Jacqueline Ulloa-Gallardo

et al.

Frontiers in Education, Journal Year: 2025, Volume and Issue: 10

Published: May 9, 2025

School principals encounter contemporary demands that impact their job satisfaction and leadership effectiveness. Despite the significance of this issue, there is limited research on predictors for these professionals, particularly using machine learning approaches. This study identified key among Peruvian school by applying an ensemble feature selection methods evaluating five algorithms (Random Forest, Decision Trees-CART, Histogram-Based Gradient Boosting, XGBoost, LightGBM) with data from 2018 National Survey Directors. The principal variables included salary, geographic location educational institution, relationships students teachers, workplace climate, student achievements, benefits. Economic factors proved important, such as gross net income, minimum monthly amount required to meet household needs. Time-related aspects also exerted influence, including hours dedicated training, time spent administrative and/or teaching duties outside working hours, travel Local Educational Management Unit (UGEL), duration stays at UGEL, commuting residence institution. Boosting algorithm, optimized Bayesian techniques trained balanced through Random Oversampling, achieved a accuracy 0.63 test set real-world class distribution. When Generative Adversarial Networks balance only training set, better results were obtained in recall (0.74), precision (0.72), F1 score (0.70). SHAP analysis revealed economic primarily influenced dissatisfied principals, while interpersonal more important highly satisfied suggesting hierarchical pattern findings could inform strategies enhance principals' strengthen data-driven policies.

Language: Английский

A machine learning approach to identifying key predictors of Peruvian school principals' job satisfaction DOI Creative Commons
Luis Alberto Holgado-Apaza, Dany Dorian Isuiza-Pérez, Nelly Jacqueline Ulloa-Gallardo

et al.

Frontiers in Education, Journal Year: 2025, Volume and Issue: 10

Published: May 9, 2025

School principals encounter contemporary demands that impact their job satisfaction and leadership effectiveness. Despite the significance of this issue, there is limited research on predictors for these professionals, particularly using machine learning approaches. This study identified key among Peruvian school by applying an ensemble feature selection methods evaluating five algorithms (Random Forest, Decision Trees-CART, Histogram-Based Gradient Boosting, XGBoost, LightGBM) with data from 2018 National Survey Directors. The principal variables included salary, geographic location educational institution, relationships students teachers, workplace climate, student achievements, benefits. Economic factors proved important, such as gross net income, minimum monthly amount required to meet household needs. Time-related aspects also exerted influence, including hours dedicated training, time spent administrative and/or teaching duties outside working hours, travel Local Educational Management Unit (UGEL), duration stays at UGEL, commuting residence institution. Boosting algorithm, optimized Bayesian techniques trained balanced through Random Oversampling, achieved a accuracy 0.63 test set real-world class distribution. When Generative Adversarial Networks balance only training set, better results were obtained in recall (0.74), precision (0.72), F1 score (0.70). SHAP analysis revealed economic primarily influenced dissatisfied principals, while interpersonal more important highly satisfied suggesting hierarchical pattern findings could inform strategies enhance principals' strengthen data-driven policies.

Language: Английский

Citations

0