PREDICTING STUDENT PERFORMANCE IN ONLINE ENGLISH LANGUAGE LEARNING DURING CHALLENGING TIMES THROUGH LEARNING ANALYTICS

Date

2025-3-03

Author

Çelikbağ, Mehmet Ali

Metadata

Show full item record

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Item Usage Stats

51
views

16
downloads

Cite This

Learning a foreign language is a complex and significant process that influences academic, business, and social life. The COVID-19 pandemic profoundly impacted education, making online English language learning a new and challenging experience for students. This shift raised important research questions about factors affecting students’ academic performance. Under normal circumstances, university students would begin their language studies face-to-face, but the pandemic necessitated online learning. The study involved 481 students from diverse backgrounds and departments. Various features influenced academic performance at different levels and across language skills, including use of language, writing, and speaking. Gateway and proficiency exams were analyzed to develop a comprehensive understanding of online English language learning during the pandemic. The number of logins, assignment submissions, and attendance in virtual classrooms often played a crucial role in predicting achievement in online English language learning. To balance data distribution, SMOTE was applied as an oversampling technique, and 10-fold stratified sampling was used to reduce sampling bias. Several classification algorithms were tested, with Logistic Regression and Naïve Bayes performing well in most cases. Additionally, Gradient Boosting, Neural Networks, Random Forest, and SVM were effective in predicting student achievement. The findings highlight the role of instructional design and technology in facilitating online learning, particularly in times of crisis. Learning analytics considerations and further implications were explored to enhance English language education in higher education. By leveraging technology and data-driven approaches, universities can optimize online learning experiences and better support students in achieving academic success.

Subject Keywords

Online English Language Learning, Learning Analytics, Predicting Student Performance, Online Learning During the COVID-19 Pandemic

URI

https://hdl.handle.net/11511/114112

Collections

Graduate School of Natural and Applied Sciences, Thesis