Scholarlink

<p>In many universities and educational institutions around the world, the failure in the performance of the students in certain subjects <i>–dropping out or not passing–</i> is an issue of great concern. One of the problems inherent to this issue is how and when to identify students at risk in order to act accordingly, allowing for more effective and personalized interventions. In this context, the main aim of this work is to identify patterns in the performance of the students in two subjects taught during the first academic year –<i>Introduction to algebra</i> and <i>Introduction to calculus</i>–, using data from 2,292 students, collected between 2018 and 2022, applying to the engineering degree at the University of Tarapacá (Chile). Clustering techniques are used to determine the clustering model that best reflects the behaviour of the students. Specifically, ten clustering techniques with their optimal configuration have been used and evaluated, by combining some internal validation metrics together with some association validation metrics. As a result, hidden patterns that provide teachers and educational institutions with an understanding of the profile of their students have been identified, allowing for more effective and personalized interventions in the two subjects mentioned above. It has been possible to identify models for distinguishing students who drop out or fail versus those who pass the course, and also those drop out versus those who continue with the course. The methodology used throughout the work is generalizable and applicable to other subjects, as long as the necessary educational data are available.</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

Identification of Patterns to Prevent the Failure of Engineering Students

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年7月24日

查看全文
  • Fabian Santiago-Muñoz,
  • Mikel Larrañaga,
  • Jon A. Elorriaga,
  • Ana Arruarte

摘要

In many universities and educational institutions around the world, the failure in the performance of the students in certain subjects –dropping out or not passing– is an issue of great concern. One of the problems inherent to this issue is how and when to identify students at risk in order to act accordingly, allowing for more effective and personalized interventions. In this context, the main aim of this work is to identify patterns in the performance of the students in two subjects taught during the first academic year –Introduction to algebra and Introduction to calculus–, using data from 2,292 students, collected between 2018 and 2022, applying to the engineering degree at the University of Tarapacá (Chile). Clustering techniques are used to determine the clustering model that best reflects the behaviour of the students. Specifically, ten clustering techniques with their optimal configuration have been used and evaluated, by combining some internal validation metrics together with some association validation metrics. As a result, hidden patterns that provide teachers and educational institutions with an understanding of the profile of their students have been identified, allowing for more effective and personalized interventions in the two subjects mentioned above. It has been possible to identify models for distinguishing students who drop out or fail versus those who pass the course, and also those drop out versus those who continue with the course. The methodology used throughout the work is generalizable and applicable to other subjects, as long as the necessary educational data are available.