Programlama Eği̇ti̇mi̇nde Üretken Yapay Zekânin Entegrasyonu: Öğrenci̇ Etki̇leşi̇mleri̇ Ve Öğrenme Çiktilarina Yöneli̇k Keşfedi̇ci̇ Bi̇r Çalişma
Year: 2026
Type: dissertation
Abstract: This study examines how students interact with ChatGPT in an introductory programming course and how instructional approaches shape their interaction patterns, learning outcomes, and perceptions. Using a mixed-methods design, three integration strategies (unguided use, explicit instruction on effective prompting, and structured scaffolding) were implemented across three in-class practice sessions during a semester. Quantitative data were gathered through a prior knowledge test and post-session quizzes, while qualitative data consisted of ChatGPT interaction logs and student reflections. A total of 158 undergraduate students participated, and 2,388 ChatGPT prompts were thematically analyzed. Based on prompting behaviors, different profiles of interaction were identified such as “AI-Reliant Code Generators”, “AI-Reliant Code Generator & Refiners”, “AI-Assisted Code Refiners”, “AI-Collaborative Coders”, and “AI-Independent Coders”. Statistical analyses showed significant differences in profile distributions across instructional conditions, demonstrating that pedagogical design strongly influences how students engage with AI tools. Students with lower prior programming knowledge were more likely to depend on ChatGPT for direct solutions, whereas those with stronger backgrounds tended to use the tool reflectively for revision and improvement. Post-test results indicated a significant performance improvement following explicit instruction, highlighting the value of prompt-writing training and AI literacy. Qualitative findings revealed a shift from emphasizing accessibility and convenience toward valuing learning support, task progress, and personalized assistance. Despite these gains, challenges such as misinterpretation, limited feedback, and overreliance persisted. Overall, the findings underscore the importance of instructional mediation in AI-enhanced programming education.
Author: güner, hacer
Language: English
Subfield: Computer Science Applications
Field: Computer Science
Domain: Physical Sciences
SDG: Quality education
Open Access status: green
OpenAlex ID: W7128287385