Moodle interactions and academic performance: educational data mining in a Philippine university

Journal of Education and Learning (EduLearn) Vol. 19, No. 1, February 2025, pp. 542~550 ISSN: 2089-9823 DOI: 10.11591/edulearn.v19i1.21549  542 Moodle interactions and academic performance: educational data mining in a Philippine university Jamal Kay B. Rogers1,3, Tamara Cher R. Mercado1, Ronald S. Decano2,3 1 College of Information and Computing, Faculty of Information Technology and Computer Science, University of Southeastern Philippines, Davao City, Philippines 2 Institute of Teacher Education, Faculty of Teacher Education, Davao del Norte State College, Panabo City, Philippines 3 Department of Graduate School, University of the Immaculate Conception, Davao City, Philippines Article Info ABSTRACT Article history: Poor academic performance remains among the most concerning educational issues, especially in higher education and online learning. To address the concern, institutions like the University of Southeastern Philippines (USeP) leverage educational data mining (EDM) techniques to generate relevant information from learning management systems (LMS) like Moodle, supporting the overall student learning experience. Moodle, considered the most widely used LMS platform, allows researchers and educators to access course logs to generate valuable insights. This EDM study at USeP explored the relationship between Moodle interactions and academic performance using data wrangling and correlation analysis. By examining various interactions from 16 courses collected with a sample size of 682, the study revealed weak correlations between students' Assignment, Create, and Forum actions and academic performance. While Assignment and Create actions show a weak positive association, Forum actions exhibit a weak negative correlation. The majority of Moodle interactions demonstrate a negligible relationship with academic performance. These findings aim to inform educators and administrators about optimizing the use of Moodle to foster a supportive digital learning environment at USeP. This study recommends further explorations, analyses, and other approaches to deepen understanding of the relationship between Moodle interactions and academic performance. Received Dec 5, 2023 Revised Apr 1, 2024 Accepted May 18, 2024 Keywords: Association Correlation Data wrangling Feature engineering Learning management system Online learning Relationship This is an open access article under the CC BY-SA license. Corresponding Author: Jamal Kay B. Rogers College of Information and Computing, Faculty of Information Technology and Computer Science University of Southeastern Philippines Iñigo Street, Obrero, Davao City, Philippines Email: 1. INTRODUCTION Poor academic performance is one of the major concerns in educational institutions worldwide. Poor academic performance leads to students failing to achieve the minimum requirements of enrolled courses, resulting in dropouts [1]. The dropout concern is prevalent in higher education, offering online or blended courses [2]. The dropout rate for online courses worldwide is 25% to 90%, significantly higher than traditional face-to-face courses [3]. In the Philippines, the Universal Access to Quality Tertiary Education Act, an act providing free tuition and other school fees to State Universities and Colleges (SUCs), Local Universities and Colleges (LUCs), and State-Run Technical-Vocational Institutions (STVIs), has resulted to Journal homepage: http://edulearn.intelektual.org J Edu & Learn ISSN: 2089-9823  543 increased student enrollments. However, this increase has also led to an alarming overall dropout rate of 83.7% at some point in the country, regardless of learning modality [4]. In recent years, institutions have leveraged technological advancements to support students' overall learning experience. Institutions leverage the power of educational data mining (EDM) to cope with educational challenges. Using data mining techniques, EDM extracts valuable insights from educational data [5]. EDM has increased research interest over the years [6] due to its practical and advanced approach to analyzing data. The popularity of online learning through the use of learning management systems (LMS) has dramatically enhanced EDM opportunities due to the vast amounts of data that can be generated [7], [8]. Academic institutions use LMS, one of the most popular online learning platforms, to deliver online, flexible, and blended learning modalities. LMS is progressing in higher educational institutions in developed and developing countries [9] due to its many advantages, such as distance learning, automated grading, and data storage. With LMS comes a considerable amount of data users produce, and a research trend continues to grow involving analysis of student interaction and learning analytics within the LMS [10]. LMS includes various platforms such as Blackboard, Google Classroom, Canvas, and Moodle. Moodle, an open-source LMS, is the most popular among these LMS platforms, with over 160,000 registered sites worldwide, according to the 2023 Moodle site registration statistics. Moodle has gained research interest over the years, becoming the most researched LMS based on the number of SSCI-index articles published in the Web of Science database [11]. In Moodle, information about how students interact with the LMS can be collected through EDM techniques. Moodle collects information from students, such as the frequency of course access and submission of requirements [12]. This information can be accessed through the Moodle database or the Moodle logs. Over the years, LMS use has been proven beneficial to students’ multiple times [13]. While other factors contribute to the success of students, such as social, human, and reinforcement factors [9], several studies [14]–[18] suggest correlations between LMS interactions and academic performance. In contrast, some studies, such as [19], found significant and non-significant relationships. Some studies also found LMS interactions to be predictors of student success. However, these studies do not generalize findings since LMS course designs and the variables used in the analysis differ from course to course [19]. Therefore, there is a need to institutionalize the research involving LMS to generate meaningful results for an institution. Higher learning institutions in the Philippines use LMS to deliver online or blended courses. The University of the Philippines (UP), De La Salle University, and the University of Sto. Thomas integrated LMS into their instructional offerings. However, there is still a concern about LMS acceptance, especially among students who are not ready to embrace technological advancements [20]. Nevertheless, this concern did not hinder Philippine institutions from using LMS, especially during the COVID-19 pandemic. For instance, Isabela State University developed a Moodle-based LMS as a customized LMS to fit their needs [21]. Their Moodle-based LMS has proven more effective in t (...truncated)