Measuring the Impact of Claude AI on University Students’ Recount Writing Proficiency

Mufadillah, Wilda (2026) Measuring the Impact of Claude AI on University Students’ Recount Writing Proficiency. Diploma atau S1 thesis, UIN Sultan Maulana Hasanuddin Banten.

[img] Teks
S_TBI_211230104_Cover.pdf

Download (509kB)
[img] Teks
S_TBI_211230104_Statement of Originality.pdf

Download (585kB)
[img] Teks
S_TBI_211230104_Chapter I.pdf

Download (482kB)
[img] Teks
S_TBI_211230104_Chapter II.pdf
Restricted to Hanya staf repositori

Download (398kB)
[img] Teks
S_TBI_211230104_Chapter III.pdf
Restricted to Hanya staf repositori

Download (394kB)
[img] Teks
S_TBI_211230104_Chapter IV.pdf
Restricted to Hanya staf repositori

Download (527kB)
[img] Teks
S_TBI_211230104_Chapter V.pdf

Download (214kB)
[img] Teks
S_TBI_211230104_References.pdf

Download (322kB)

Abstrak

This study investigates the effectiveness of Claude AI in improving first-semester students’ recount text writing, as well as their perceptions of using the tool during the writing process. The rapid development of AI-based writing assistants provides new opportunities for enhancing linguistic accuracy, self-editing skills, and learner autonomy in EFL contexts. Using an explanatory mixed-methods design, this research involved 20 students selected purposively from the English Language Education Study Program. Quantitative data were obtained by analyzing students’ recount texts using Claude AI to identify linguistic errors across seven categories: sentence structure, capitalization, spelling, punctuation, grammar, mechanics, and vocabulary. Qualitative data were collected through semi-structured interviews with eight participants representing high, moderate, and low proficiency levels. The quantitative results reveal that sentence structure (28.6%), capitalization (27.8%), and spelling (21.9%) were the most frequent errors prior to revision, indicating that students struggled with syntactic accuracy and basic writing mechanics. After using Claude AI, students demonstrated noticeable improvement in grammatical accuracy, writing organization, and mechanical correctness. The qualitative results further show that students perceived Claude AI as helpful, motivating, and easy to use, stating that it raised their awareness of recurring mistakes, reduced writing anxiety, and supported independent revision. However, challenges such as overreliance, general feedback, and difficulty understanding complex explanations were also reported. The study concludes that Claude AI effectively enhances students’ recount writing performance while fostering confidence and autonomy, yet requires balanced use and teacher guidance to prevent dependency and ensure long-term skill development.

Tipe Item/Data: Skripsi/Tesis/Disertasi (Diploma atau S1)
Kata Kunci (keywords): Claude AI, recount text, writing improvement, AI feedback, EFL learners, error analysis, student perception
Subjek: 400 Bahasa > 420 Bahasa Inggris & bahasa Inggris Kuno > 428 Pemakaian Bahasa Inggris Baku
Divisi: Fakultas Tarbiyah dan Keguruan > Tadris Bahasa Inggris
User Penyetor: S.S.I Fadhilah NH
Tanggal Disetorkan: 01 Jul 2026 11:58
Perubahan Terakhir: 01 Jul 2026 11:58
URI: http://repository.uinbanten.ac.id/id/eprint/18814

Actions (login required)

Lihat Item Lihat Item

      is powered by EPrints 3 which is developed by the Islamic Institutional Repository UIN Sultan Maulana Hasanuddin Banten. More information and software credits.