Computational Modeling of Teacher Performance Using Leadership Style and Professional Competence

Teacher performance is a critical indicator of educational quality, yet existing studies have largely examined leadership style, work ethic, work motivation, and professional competence independently, leaving a gap in computational approaches that model their combined influence on teacher performance. This study aims to develop a computational modeling framework for predicting teacher performance by examining the relationships among leadership style, work ethic, work motivation, and professional competence in Early Childhood Education (ECE) institutions. A quantitative explanatory approach was employed using survey data collected from 361 teachers in Medan Satria District, Bekasi City, Indonesia. The proposed structural relationships were evaluated using Structural Equation Modeling (SEM) with AMOS to estimate the predictive contributions of each construct. The results indicate that leadership style and professional competence significantly predict teacher performance, with professional competence demonstrating the strongest effect. In contrast, work ethic and work motivation exhibit statistically significant but relatively weak practical contributions, while professional competence does not mediate the relationship between leadership style and teacher performance. These findings demonstrate that computational modeling based on structural relationships can support evidence-based educational management by identifying the most influential determinants of teacher performance. The study contributes an integrated computational framework for teacher performance prediction and provides practical insights for educational decision-makers. Future research should address the current gap by incorporating machine learning, learning analytics, and explainable artificial intelligence to improve predictive accuracy and enable intelligent decision support in digital education environments.