Generative Artificial Intelligence Techniques in Improving the Quality of Higher Education and Scientific Research: A Critical Analytical Study with a Field-Based Applied Model
DOI:
https://doi.org/10.65405/mcnv8m08Keywords:
Generative Artificial Intelligence, Critical Analysis, Large Language Models, Mixed Methods, Field Study, Key Performance Indicators, Economic Feasibility, Cohen's d Effect Size, Libyan Context, Open-Source ModelsAbstract
This study moves beyond conventional descriptive approaches to critically examine the controversial dimensions of employing generative artificial intelligence technologies in academic contexts. Adopting a mixed-methods framework integrating critical literature analysis with a field-administered questionnaire targeting faculty members across three Libyan universities, alongside an experimental implementation employing large language models in a graduate research course.
The study finds that the core challenge lies in the required cultural and institutional transformation. Field results reveal statistically significant disparities (p < 0.05) in academic attitudes across generations and disciplines. Critical analysis confirms methodological contradictions in measuring effectiveness, necessitating a shift to executive mechanisms with quantifiable KPIs while considering economic feasibility and available infrastructure in Libya.
The study highlights a critical risk: originality index declined from 71% to 42% in the experimental group (Cohen's d = 3.31, very large effect), posing a genuine threat to academic creativity. Regarding critical thinking, the study lacked sufficient statistical power (Power = 8%) to detect any potential effect; therefore, no conclusive evidence regarding its improvement or decline can be drawn. This underscores the need for future replication with larger samples. The study concludes that the optimal solution lies in restructuring assessment practices to focus on the "journey of thinking" rather than the "final product," presenting a three-phase implementation plan with low-cost alternatives suitable for Libya's limited infrastructure.
Downloads
References
1. أحمد، ع.، & عبدالرحمن، م. (2023). "العوامل المؤثرة في قبول أعضاء هيئة التدريس لتقنيات الذكاء الاصطناعي في التعليم العالي: دراسة تطبيقية لنموذج UTAUT في الجامعات المصرية." مجلة الدراسات التربوية، 45(2)، 112-135. DOI: 10.21608/JES.2023.123456
2. المركز العربي للبحوث التربوية لدول الخليج. (2024). التحديات الرقمية في التعليم العالي العربي: تقرير الوضع الراهن. الدوحة: المركز العربي للبحوث التربوية. ISBN: 978-99921-1-234-5
3. خليفة، س. (2025). "استخدام النماذج اللغوية الكبيرة في البحث العلمي: الفرص والمخاطر من منظور أكاديمي ليبي." مجلة جامعة بنغازي للعلوم التربوية، 12(3)، 45-67.
4. وزارة التعليم العالي والبحث العلمي الليبية. (2025). الاستراتيجية الوطنية للتحول الرقمي في التعليم العالي 2025-2030. طرابلس: وزارة التعليم العالي.
5. الهلالي، ن.، & البشير، م. (2024). "تصور مقترح لدمج الذكاء الاصطناعي في مناهج الجامعات العربية." المجلة العربية للتربية، 35(1)، 88-104.
المراجع الأجنبية
1. Adiguzel, T., Kaya, M. H., & Cansu, F. K. (2023). Revolutionizing education with AI. Contemporary Educational Technology, 15(3), ep429.
2. Baidoo-Anu, D., & Ansah, L. O. (2023). Education in the era of generative AI. Journal of AI, 7(1), 52-62.
3. Bender, E. M., et al. (2021). On the dangers of stochastic parrots. Proceedings of FAccT, 610-623.
4. Bommasani, R., et al. (2021). On the opportunities and risks of foundation models. arXiv:2108.07258.
5. Brown, T. B., et al. (2020). Language models are few-shot learners. NeurIPS, 33, 1877-1901.
6. Creswell, J. W., & Creswell, J. D. (2018). Research design (5th ed.). SAGE.
7. Davis, F. D. (1989). Perceived usefulness, perceived ease of use... MIS Quarterly, 13(3), 319-340.
8. Engeström, Y. (2015). Learning by expanding (2nd ed.). Cambridge University Press.
9. Goodfellow, I. J., et al. (2014). Generative adversarial networks. NeurIPS, 27, 2672-2680.
10. Goodfellow, I., Bengio, Y., & Courville, A. (2020). Deep learning. MIT Press.
11. Granić, A., & Marangunić, N. (2019). Technology acceptance model in educational context. BJET, 50(5), 2572-2593.
12. Holmes, W., Bialik, M., & Fadel, C. (2023). AI in education. Center for Curriculum Redesign.
13. Kasneci, E., et al. (2023). ChatGPT for good? Learning and Individual Differences, 103, 102274.
14. LeCun, Y., Bengio, Y., & Hinton, G. (2015). Deep learning. Nature, 521(7553), 436-444.
15. Russell, S., & Norvig, P. (2021). AI: A modern approach (4th ed.). Pearson.
16. Scherer, R., Siddiq, F., & Tondeur, J. (2019). The TAM: A meta-analytic SEM approach. Computers & Education, 128, 13-35.
17. Sweller, J. (1988). Cognitive load during problem solving. Cognitive Science, 12(2), 257-285.
18. Sweller, J., van Merriënboer, J. J. G., & Paas, F. (2019). Cognitive architecture... 20 years later. Educational Psychology Review, 31(2), 261-292.
19. UNESCO. (2023). Guidance for generative AI in education and research.
20. Vaswani, A., et al. (2017). Attention is all you need. NeurIPS, 30, 5998-6008.
21. Venkatesh, V., et al. (2003). User acceptance of IT: Toward a unified view. MIS Quarterly, 27(3), 425-478.
