DOI QR코드

DOI QR Code

A Study on Predicting Student Dropout in College: The Importance of Early Academic Performance

전문대학 학생의 학업중단 예측에 관한 연구: 초기 학업 성적의 중요성

  • Sangjo Oh (Dept. of Management Information Systems, Dongyang Mirae University) ;
  • JiHwan Sim (Dept. of Big Data Management, Dongyang Mirae University)
  • 오상조 (동양미래대학교 경영정보학과) ;
  • 심지환 (동양미래대학교 빅데이터경영과)
  • Received : 2023.12.28
  • Accepted : 2024.02.20
  • Published : 2024.02.28

Abstract

This study utilized minimum number of demographic variables and first-semester GPA of students to predict the final academic status of students at a vocational college in Seoul. The results from XGBoost and LightGBM models revealed that these variables significantly impacted the prediction of students' dismissal. This suggests that early academic performance could be an important indicator of potential academic dropout. Additionally, the possibility that academic years required to award an associate degree at the vocational college could influence the final academic status was confirmed, indicating that the duration of study is a crucial factor in students' decisions to discontinue their studies. The study attempted to model without relying on psychological, social, or economic factors, focusing solely on academic achievement. This is expected to aid in the development of an early warning system for preventing academic dropout in the future.

본 연구에서는 서울 소재 한 전문대학 학생들을 대상으로 하여 최소한의 인구통계학적 변수와 1학년 1학기 성적을 활용하여 학생들의 최종 학적 상태를 예측하고자 하였다. XGBoost와 LightGBM 모델을 사용한 결과, 이러한 변수들이 학생들의 제적 여부 예측에 유의미한 것을 발견하였다. 이는 학업 시작 초기의 성적이 학업 중단의 중요한 지표가 될 수 있음을 시사한다. 또한, 전문대학의 학제가 최종 학적에 영향을 미칠 가능성을 확인하였으며, 이는 학업 기간이 학생들의 학업 중단 결정에 중요한 요소임을 나타낸다. 전문대학에서 조기 학업 중단 의도를 파악하는 데 있어 심리적, 사회적, 경제적 요인에 의존하지 않고 학업 성취도만을 기준으로 모델링을 시도하였다. 이는 향후 학업 중단에 대한 조기 경보 시스템 구축에 도움이 될 것으로 기대된다.

Keywords

Acknowledgement

This paper was supported by Dongyang Mirae University research fund in 2023.

References

  1. Agrusti, F., Bonavolonta, G., & Mezzini, M. (2019). University dropout prediction through educational data mining techniques: A systematic review. Journal of e-learning and knowledge society, 15(3), 161-182. DOI : 10.20368/1971-8829/1135017
  2. Aina, C., Baici, E., Casalone, G., & Pastore, F. (2022). The determinants of university dropout: A review of the socio-economic literature. Socio-Economic Planning Sciences, 79, 101102. DOI : 10.1016/j.seps.2021.101102
  3. Alban, M., & Mauricio, D. (2019). Predicting university dropout through data mining: a systematic literature. Indian Journal of Science and Technology, 12(4), 1-12. DOI : 10.17485/ijst/2019/v12i4/139729
  4. Ameen, A. O., Alarape, M. A., & Adewole, K. S. (2019). Students' academic performance and dropout predictions: A review. Malaysian Journal of Computing, 4(2), 278-303. DOI : 10.24191/mjoc.v4i2.6701
  5. Astin, A. W. (1964). Personal and environmental factors associated with college dropouts among high aptitude students. Journal of Educational Psychology, 55(4), 219. DOI : 10.1037/h0046924
  6. Aulck, L., Velagapudi, N., Blumenstock, J., & West, J. (2016). Predicting student dropout in higher education. arXiv preprint arXiv:1606. 06364. DOI : 10.48550/arXiv.1606.06364
  7. Behr, A., Giese, M., Teguim K, H. D., & Theune, K. (2020). Early prediction of university dropoutsa random forest approach. Jahrbucher fur Nationalokonomie und Statistik, 240(6), 743-789. DOI : 0.1515/jbnst-2019-0006 https://doi.org/10.1515/jbnst-2019-0006
  8. Bernardo, A. B., Galve-Gonzalez, C., Nunez, J. C., & Almeida, L. S. (2022). A path model of university dropout predictors: the role of satisfaction, the use of self-regulation learning strategies and students' engagement. Sustainability, 14(3), 1057.
  9. Del Bonifro, F., Gabbrielli, M., Lisanti, G., & Zingaro, S. P. (2020). Student dropout prediction. In Artificial Intelligence in Education: 21st International Conference, AIED 2020, Ifrane, Morocco, July 6-10, 2020, Proceedings, Part I 21, 129-140. DOI : 10.3390/su14031057
  10. Delen, D. (2010). A comparative analysis of machine learning techniques for student retention management. Decision Support Systems, 49(4), 498-506. DOI : 10.1016/j.dss.2010.06.003
  11. Demeter, E., Dorodchi, M., Al-Hossami, E., Benedict, A., Slattery Walker, L., & Smail, J. (2022). Predicting first-time-in-college students' degree completion outcomes. Higher Education, 1-21. DOI : 10.1007/s10734-021-00790-9
  12. Lee, S., & Chung, J. Y. (2019). The machine learning-based dropout early warning system for improving the performance of dropout prediction. Applied Sciences, 9(15), 3093. DOI : 10.3390/app9153093
  13. Lizarte Simon, E. J., & Gijon Puerta, J. (2022). Prediction of early dropout in higher education using the SCPQ. Cogent Psychology, 9(1), 2123588. DOI : 10.1080/23311908.2022.2123588
  14. Lounsbury, J. W., Saudargas, R. A., & Gibson, L. W. (2004). An investigation of personality traits in relation to intention to withdraw from college. Journal of College Student Development, 45(5), 517-534. DOI : 10.1353/csd.2004.0059
  15. Lykourentzou, I., Giannoukos, I., Nikolopoulos, V., Mpardis, G., & Loumos, V. (2009). Dropout prediction in e-learning courses through the combination of machine learning techniques. Computers & Education, 53(3), 950-965. DOI : 10.1016/j.compedu.2009.05.010
  16. Mduma, N., Kalegele, K., & Machuve, D. (2019). A survey of machine learning approaches and techniques for student dropout prediction. Data Science Journal, 18, 14-14.
  17. Mortada, L., Bolbol, J., & Kadry, S. (2018). Factors affecting students' performance a case of private colleges in Lebanon. J Math Stat Anal, 1, 105. DOI : 10.5334/dsj-2019-014
  18. Niyogisubizo, J., Liao, L., Nziyumva, E., Murwanashyaka, E., & Nshimyumukiza, P. C. (2022). Predicting student's dropout in university classes using two-layer ensemble machine learning approach: A novel stacked generalization. Computers and Education: Artificial Intelligence, 3, 100066. DOI : 10.1016/j.caeai.2022.100066
  19. OECD. (2019). Education at a Glance 2019 : OECD Indicators. Paris : OECD Publishing. DOI : 10.1787/19991487
  20. Orooji, M., & Chen, J. (2019, December). Predicting louisiana public high school dropout through imbalanced learning techniques. In 2019 18th IEEE International Conference On Machine Learning And Applications (ICMLA) (pp. 456-461). IEEE. DOI : 10.1109/ICMLA.2019.00085
  21. Shafiq, D. A., Marjani, M., Habeeb, R. A. A., & Asirvatham, D. (2022). Student retention using educational data mining and predictive analytics: a systematic literature review. IEEE Access. DOI : 10.1109/ACCESS.2022.3188767.
  22. Tan, M., & Shao, P. (2015). Prediction of student dropout in e-Learning program through the use of machine learning method. International journal of emerging technologies in learning, 10(1). DOI : 10.3991/ijet.v10i1.4189
  23. Thammasiri, D., Delen, D., Meesad, P., & Kasap, N. (2014). A critical assessment of imbalanced class distribution problem: The case of predicting freshmen student attrition. Expert Systems with Applications, 41(2), 321-330. DOI : 10.1016/j.eswa.2013.07.046
  24. Tinto, V. (1975). Dropout from higher education: A theoretical synthesis of recent research. Review of educational research, 45(1), 89-125. DOI : 10.2307/1170024
  25. Tinto, V. (2006). Research and practice of student retention: What next?. Journal of college student retention: Research, Theory & Practice, 8(1), 1-19. DOI : 10.2190/4YNU-4TMB-22DJ-AN4W
  26. Young A, Song., Sinae, Kim. (2019). Factors Affecting College Freshmen's Intention to Drop Out. The Korea Contents Association, 19(6), 257-270. DOI : 10.5392/JKCA.2019.19.06.257
  27. Youngsik, Woo., Minok, Song. (2022). Relationship Between Carrer Decision Level, Academic Self-effcacy, Self-directed Learning Ability, and College Life Adptation of Junior College Freshmen. The Journal of Humanities and Social Science 21, 13(4), 1417-1432.
  28. Kostat. (2023). Birth Statistics in 2022 [Press Release].