Modeling the Impact of Engagement Parameters on Student Performance in Blended Learning by using Machine Learning | IJCT Volume 13 – Issue 1 | IJCT-V13I1P2
Modeling the Impact of Engagement Parameters on Student Performance in Blended Learning by using Machine Learning | IJCT Volume 13 – Issue 1 | IJCT-V13I1P2
The expeditious growth of blended learning has created a demand for analytical methods that can quantify how digital engagement affects educational outcomes across various academic programs. This study will reveal a cross-course evaluation that combine supervised machine learning, engagement Modeling and comparative performance analysis to better understand the pupil’s behaviour within blended and traditional instructional conditions. By using the dataset partitions representing four academic programs- B.Tech 2nd year, B.Tech 3rd year, B.Sc 3rd year and BCA 3rd year-the study examines the relationship between structured engagement in blended learned subjects and resulting academic performance. The analysis comprise performance averages, engagement-to-performance correlations, and model-driven predictions created through ML algorithms such as Linear Regression, Decision Trees, Random Forests, and Support Vector Regression. Findings show variations in blended learning effectiveness across these courses, with engagement demonstrating powerful predictive value in computer engineering programs and modest or inconsistent alliance in other disciplines. The recent work emphasizing learner analytics and AI-supported evaluation setup in higher education. The study highlights how engagement parameters and algorithmic modeling can support an extensive understanding of blended learning dynamics, offering evidence-based understanding for instructional design, policy decisions, and future adaptive learning research.
Keywords
Machine Learning, Model Prediction, Blended Learning, Parameter Influence, Performance Analysis.
Conclusion
The findings of this study highlight the complex and program-dependent nature of blended learning effectiveness within higher education environments. By analysing four distinct academic cohorts-spanning engineering, computer applications, and science-clear differences emerged in how students responded to the blended instructional mode. These differences reflect not only variation in student preparedness and subject complexity but also the structural alignment between digital learning components and learners’ academic trajectories. Similar patterns have been noted in other cross-disciplinary evaluations, where blended learning demonstrated uneven performance benefits across domains.
The engineering cohorts (B.Tech 2nd and 3rd year) demonstrated the most favourable response to blended instruction. These programs exhibited both strong blended performance and high positive performance gains relative to traditional subjects. The correlation analysis reinforced these outcomes, particularly in the B.Tech 3rd year group, where engagement was strongly associated with improved blended performance. This finding suggests that when digital components are embedded in conceptually structured courses that require visualisation, modelling, or iterative exploration, students may benefit substantially from the additional digital exposure. The predictive modelling results further supported this interpretation, with high R² values indicating that engagement was a reliable predictor of blended performance in these groups. These outcomes align with recent work that demonstrates the importance of alignment between digital task design and subject-specific cognitive requirements.
In contrast, the B.Sc 3rd year cohort presented a markedly different pattern. Despite exhibiting one of the highest engagement-to-performance correlations, the group underperformed in blended settings relative to their traditional subjects, yielding a negative performance gain. This discrepancy suggests that although engagement was consistent, it did not translate into academic benefit. The divergence between blended and traditional outcomes may signal limitations in the adaptability of the blended content to the students’ domain knowledge or learning preferences. Factors such as digital literacy, subject complexity, or cognitive load may have constrained students from maximising the benefits of the blended format-an observation consistent with studies indicating that blended learning can pose challenges when learners are insufficiently prepared for self-regulated digital study.
The BCA 3rd year cohort showed a neutral effect, with nearly identical blended and traditional averages. Models trained on the BCA data performed poorly, as indicated by negative R² scores. This is likely due to the lack of variance in engagement-every student received the same engagement value-and the closely clustered distribution of performance scores. When both the predictor and outcome variables show minimal dispersion, supervised models have limited capacity to detect meaningful trends. This reinforces the principle that predictive learning analytics are most effective when engagement measures reflect actual behavioural differences, rather than uniform instructional inputs.
Taken together, the cross-cohort results reveal that blended learning effectiveness depends not only on the presence of digital components but also on the degree of meaningful learner engagement, program-specific instructional design, and cognitive alignment. Programs where digital elements complement the disciplinary learning structure tend to benefit more, while programs requiring closer instructor mediation or sequential scaffolding may not exhibit similar advantages. This supports the broader argument that blended learning cannot be designed uniformly across disciplines; its outcomes are inherently shaped by pedagogical context and learner readiness.
References
[1]Asiksoy, G. (2025). Predicting Student Performance in a Blended Learning Course Using Machine Learning. In Smart Infrastructures in the IoT Era (pp. 1251-1263). Cham: Springer Nature Switzerland.
[2]Nguyen, H. T., Pham, L. T., Nguyen, V. A., & Do, K. T. (2025). Early prediction of student performance based on behavioral data in blended learning. The International Journal of Information and Learning Technology, 42(3), 311-328.
[3]Azizah, Z., Ohyama, T., Zhao, X., Ohkawa, Y., & Mitsuishi, T. (2024). Predicting at-risk students in the early stage of a blended learning course via machine learning using limited data. Computers and Education: Artificial Intelligence, 7, 100261.
[4]T. M. Osaili, L. C. Ismail, H. M. ElMehdi, A. A. Al-Nabulsi, A. O. Taybeh, S. T. Saleh, H. Kassem, H. Alkhalidy, H. I. Ali, A. S. Al Dhaheri, and L. Stojanovska, (Mar. 2023) ‘‘Comparison of students’ perceptions of online and hybrid learning modalities during the COVID-19 pandemic: The case of the University of Sharjah,’’ PLoS ONE, vol. 18, no. 3, Art. no. e0283513, doi: 10.1371/journal.pone.0283513.
[5]Hamadneh, N. N., Atawneh, S., Khan, W. A., Almejalli, K. A., & Alhomoud, A. (2022). Using Artificial Intelligence to Predict Students’ Academic Performance in Blended Learning. Sustainability 2022, 14, 11642.
[6]Ciolacu, M. I., & Svasta, P. (2021, April). Education 4.0: AI empowers smart blended learning process with biofeedback. In 2021 IEEE Global Engineering Education Conference (EDUCON) (pp. 1443-1448). IEEE.
[7]Han, F., & Ellis, R. A. (2021). Predicting students’ academic performance by their online learning patterns in a blended course. Educational Technology & Society, 24(1), 191-204.
[8]Evangelista, E. (2021). A hybrid machine learning framework for predicting students’ performance in virtual learning environment. International Journal of Emerging Technologies in Learning (iJET), 16(24), 255-272.
[9]Han, F., & Ellis, R. A. (2021). Predicting students’ academic performance by their online learning patterns in a blended course. Educational Technology & Society, 24(1), 191-204.
[10]Alraddadi, S., Alseady, S., & Almotiri, S. (2021, March). Prediction of students academic performance utilizing hybrid teaching-learning based feature selection and machine learning models. In 2021 International Conference of Women in Data Science at Taif University (WiDSTaif) (pp. 1-6). IEEE.
[11]Xu, Z., Yuan, H., & Liu, Q. (2020). Student performance prediction based on blended learning. IEEE Transactions on Education, 64(1), 66-73.
[12]Salas-Rueda, R. A. (2020). Perception of students on blended learning considering data science and machine learning. Campus Virtuales, 9(1), 125-135.
[13]Van Goidsenhoven, S., Bogdanova, D., Deeva, G., Broucke, S. V., De Weerdt, J., & Snoeck, M. (2020, March). Predicting student success in a blended learning environment. In Proceedings of the tenth international conference on learning analytics & knowledge (pp. 17-25).
[14]Chango, W., Cerezo, R., & Romero, C. (2019, December). Predicting academic performance of university students from multi-sources data in blended learning. In Proceedings of the second international conference on data science, E-learning and information systems (pp. 1-5).
[15]Raga, R. C., & Raga, J. D. (2019, July). Early prediction of student performance in blended learning courses using deep neural networks. In 2019 International Symposium on Educational Technology (ISET) (pp. 39-43). IEEE.
[16]S. Makridakis, ‘‘The forthcoming artificial intelligence (AI) revolution: Its impact on society and firms,’’ Futures, vol. 90, pp. 46–60, Jun. 2017, doi: 10.1016/j.futures.2017.03.006.
[17]Zacharis, N. Z. (2016). Predicting student academic performance in blended learning using artificial neural networks. International Journal of Artificial Intelligence and Applications, 7(5), 17-29.
[18]S. K. Howard, J. Ma, and J. Yang, (Oct 2016) ‘‘Student rules: Exploring patterns of students’ computer-efficacy and engagement with digital technologies in learning,’’ Comput. Educ., vol. 101, pp. 29–42. doi:10.1016/j.compedu.2016.05.008.
How to Cite This Paper
Kuldeep Chauhan, Varun Bansal, Anil Kumar, Vishal Kumar (2025). Modeling the Impact of Engagement Parameters on Student Performance in Blended Learning by using Machine Learning. International Journal of Computer Techniques, 12(6). ISSN: 2394-2231.
