Exploring university students’ perspectives’ on ChatGPT integration in education
-
Lawrence Ibeh ; Noah Mutai [1] ; Priyadarshini Pattanaik ; Nguyen Manh-Cuong ; Benjamin Bensam-Sambiri [1] ; Kaddour Chelabi [1] ; Olufunke Mercy-Popoola ; Alexander Makarov ; Navya Gubbi-Sateeshchandra [1]
-
[1] Berlin School of Business and Innovation (Germany)
-
- Localización: JOTSE, ISSN-e 2013-6374, Vol. 15, Nº. 2, 2025, págs. 289-301
- Idioma: inglés
- Enlaces
-
Resumen
This study explores university students’ perceptions of ChatGPT, focusing on its educational benefits, drawbacks, and possible solutions. Data were collected through an online questionnaire completed by 350 students. A descriptive research design and Multinomial Logistic Regression (MLR) were used to analyze responses. Findings show that 53.14% believe ChatGPT positively impacts academic performance, 47.14% find it useful for writing assignments, and 50% use it for exam preparation. However, students’ express concerns about its accuracy (61.72%), reliability (52.29%), privacy risks (52.57%), potential bias (47.33%), and misuse (43.71%). Broader concerns include security vulnerabilities (55.14%) and fears about AI replacing human labour (56.29%). Recommendations include integrating AI education into curricula to help students understand both capabilities and limitations of models like ChatGPT and emphasizing ethical considerations. Educator training should focus on evaluating information reliability, recognizing bias, and critically assessing AI-generated content. Educators should also prioritize human sources when appropriate. For policymakers, enforcing regulations that uphold accuracy, reliability, privacy, and ethical standards in AI is crucial. Developers are encouraged to improve model transparency and explainability to foster trust and accountability.
