فاعلية نماذج الذكاء الاصطناعي التوليدي في تحسين مهارات الكتابة الأكاديمية لدى طلبة الجامعات

Recent years have witnessed rapid advances in generative artificial intelligence (AI) models such as ChatGPT and Gemini, opening new prospects for supporting the educational process. This study aims to measure the effectiveness of generative AI models in improving university students' academic writing skills, using a quasi-experimental design with experimental and control groups (N = 60 second-year students of the Faculty of Education, randomly assigned to the two groups). The experimental group received a six-week training program that integrated generative AI as a writing assistant across brainstorming, drafting, language revision, and referencing, while the control group followed traditional instruction. An academic writing skills test and an observation checklist were used as instruments. The study hypothesizes statistically significant differences (α = 0.05) favoring the experimental group in thesis clarity, idea organization, linguistic accuracy, and academic referencing, alongside positive attitudes toward AI as a supportive tool rather than a substitute for thinking. In addition, an exploratory descriptive survey of 87 university students found that 93.1% had previously used generative AI tools for academic writing, 95.4% supported clear university policies governing their use, and 72.4% reported concern about over-reliance on them. The study recommends training faculty members to integrate generative AI tools into academic writing courses and establishing an ethical charter to govern their use.