AI-assisted Error Analysis to Design a Writing Course Syllabus
International Conference on Computer Sciences (IHCONCS 2026), Zagreb, Croatia, Zagreb, Hırvatistan, 17 - 18 Eylül 2026, ss.1516-1525, (Tam Metin Bildiri)
- Yayın Türü: Bildiri / Tam Metin Bildiri
- Basıldığı Şehir: Zagreb
- Basıldığı Ülke: Hırvatistan
- Sayfa Sayıları: ss.1516-1525
- Açık Arşiv Koleksiyonu: AVESİS Açık Erişim Koleksiyonu
- Gazi Üniversitesi Adresli: Evet
Özet
Abstract
This study aims to identify students’ recurring writing errors through AI-assisted analysis and use the findings to design a writing course syllabus. To define course objectives, it is essential to explore students' needs. To design the syllabus for a writing course, 50 essays written by engineering students are analyzed by using GenAI technology (GPT-5.6), and the errors are identified under a pre-defined 18-category taxonomy. The findings indicate that the major language knowledge flaw lies in lexico-grammatical accuracy, particularly in collocation and subject-verb agreement, rather than in surface-level writing mechanics such as spelling or punctuation. The study recommends that the suggested course syllabus focus primarily on vocabulary, with an emphasis on teaching and learning vocabulary in chunks.
Keywords: syllabus design, error analysis, error frequency, Artificial Intelligence