Journal Home
Login
Register
Open Access Journal
Submit a Paper
Propose a Special lssue
Open Access
ARTICLE
Predictive Modeling of Programming Anxiety in University Students: A Logistic Regression Approach for Early Identification
Issue Vol. 2 No. 02 (2025): Volume 02 Issue 02 --- Section Articles --- Published Date: 2025-12-24
Abstract
The proliferation of programming as a fundamental skill across academic and professional domains has amplified the need to address pedagogical challenges that hinder student success. Among these, programming anxiety—a specific form of situational anxiety characterized by fear, apprehension, and cognitive interference during programming tasks—has emerged as a significant barrier to learning and retention in computing education. This study addresses the critical need for scalable, automated methods to identify students at risk of experiencing high levels of programming anxiety. We developed and evaluated a machine learning classification model using a dataset of 1,732 undergraduate students from computing-related degree programs. The study was structured using the Cross-Industry Standard Process for Data Mining (CRISP-DM) framework, encompassing comprehensive data preprocessing, feature engineering, and class imbalance mitigation through the Synthetic Minority Over-sampling Technique (SMOTE). Five supervised learning algorithms were systematically compared: Logistic Regression, Support Vector Machine, Naïve Bayes, Random Forest, and a Decision Tree classifier. Performance was assessed using a suite of metrics, including accuracy, precision, recall, F-measure, and Cohen's kappa, with model robustness confirmed via stratified 10-fold cross-validation. The Logistic Regression model demonstrated superior performance, achieving an accuracy of 97.75%, a precision of 96.88%, a recall of 96.70%, an F-measure of 96.77%, and a Cohen's kappa of 0.950. Key predictors identified by the model included previous academic performance in foundational programming courses, high school academic track, working student status, and sleep patterns. The resulting model provides a highly accurate and interpretable tool for educational institutions. Its potential for integration into learning management systems and academic advising platforms offers a proactive mechanism for delivering timely, targeted interventions, thereby fostering a more supportive learning environment and enhancing student success in the digital age.
Keywords
References
[1] R. Yazdanian, R. L. Davis, X. Guo, F. Lim, P. Dillenbourg, and M.-Y. Kan, 'On the radar: Predicting nearfuture surges in skills’ hiring demand to provide early warning to educators', Computers and Education: Artificial Intelligence, vol. 3, p. 100043, Jan. 2022. doi.org/10.1016/j.caeai.2021.100043
[2] K. M. N. Rebuta, I. M. P. Cabaron, R. J. C. Pucong, J. M. C. Bisquera, R. T. Llerado, and M. V. M. Buladaco, 'Relationship of programming skills and perceived value of learning programming among Information Technology education students in Davao Del Sur', Int. J. Res. Innov. Soc. Sci., vol. 6, no. 6, pp. 882–887, 2022. doi.org/10.47772/ijriss.2022.6633
[3] Philippine Commission on Higher Education, CMO No. 25, Series of 2015: Policies, standards, and guidelines for the Bachelor of Science in Information Technology program, 2015. [Online]. Available: https://ched.gov.ph/wp-content/uploads/2017/10/CMO-no.-25-s.-2015.pdf
[4] Commission on Higher Education, "CHED Memorandum Order No. 24, series of 2015: Policies, Standards and Guidelines for the Bachelor of Library and Information Science (BLIS) Program," Oct. 12, 2015. [Online]. Available: https://ched.gov.ph/wp-content/uploads/2017/10/CMO-no.-24-s.-2015.pdf
[5] L. Hu, 'Programming and 21st century skill development in K‐12 schools: A multidimensional meta‐analysis', Journal of Computer Assisted Learning, vol. 40, no. 2, pp. 610–636, Nov. 2023. doi.org/10.1111/jcal.12904
pdf
