Scholarlink

<p>The main objective of this research is to build a predictive model for evaluating student performance on a national scale in primary education settings, using Learning Management System (LMS) information. Utilizing a comprehensive dataset from primary school at national level, this study employs Bayesian Additive Regression Trees (BART) with random effects to examine the relationship between student performance, (LMS) engagement, and socioeconomic status. The methodological contribution lies in the use of BART as a flexible nonparametric approach capable of modeling complex, high-dimensional relationships while incorporating school-level heterogeneity through random effects. Furthermore, its Bayesian formulation provides a natural mechanism for quantifying predictive uncertainty. To address the difficulty of summarizing the global influence of numerous LMS activity indicators, we propose a synthetic student profile methodology that allows an interpretable assessment of the overall effect of digital engagement on learning outcomes. The findings indicate that the model is effective for the early identification of students at risk and for identifying schools that excel or require intervention. Notably, the analysis reveals that higher levels of LMS usage have a positive association with student success, especially for students from schools with lower socioeconomic contexts. These findings highlight the potential of data-driven approaches to support educational policy and resource allocation.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Student performance and LMS activity in primary schools: a Bayesian additive regression trees approach with random effects

  • 期刊论文

  • 发表日期：

    2026年4月16日

查看全文
  • Natalia da Silva,
  • Bruno Tancredi,
  • Ignacio Alvarez-Castro

摘要

The main objective of this research is to build a predictive model for evaluating student performance on a national scale in primary education settings, using Learning Management System (LMS) information. Utilizing a comprehensive dataset from primary school at national level, this study employs Bayesian Additive Regression Trees (BART) with random effects to examine the relationship between student performance, (LMS) engagement, and socioeconomic status. The methodological contribution lies in the use of BART as a flexible nonparametric approach capable of modeling complex, high-dimensional relationships while incorporating school-level heterogeneity through random effects. Furthermore, its Bayesian formulation provides a natural mechanism for quantifying predictive uncertainty. To address the difficulty of summarizing the global influence of numerous LMS activity indicators, we propose a synthetic student profile methodology that allows an interpretable assessment of the overall effect of digital engagement on learning outcomes. The findings indicate that the model is effective for the early identification of students at risk and for identifying schools that excel or require intervention. Notably, the analysis reveals that higher levels of LMS usage have a positive association with student success, especially for students from schools with lower socioeconomic contexts. These findings highlight the potential of data-driven approaches to support educational policy and resource allocation.