EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC 4.0

Fostering EFL Oral Fluency through Lightweight AI Feedback in Literature-Based Speaking Tasks: A Quasi-Experimental Study in Chinese Higher Education

Cao Xue, Putu Kerti Nitiasih, Ni Nyoman Padmadewi, Ni Wayan Surya Mahayanti

PPSDP International Journal of Education · 31.12.2025 · 4 (2) · 1673-1690

DOI: 10.59175/pijed.v4i2.797 ↗

Kurz gefasst

Fachartikel in der Zeitschrift PPSDP International Journal of Education, online erschienen am 31.12.2025. Verfasst von Cao Xue und 3 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This study investigates the effectiveness of lightweight Artificial Intelligence (AI) feedback in enhancing oral fluency among English as a Foreign Language (EFL) learners in Chinese higher education. Guided by skill acquisition theory, a quasi-experimental design with intact classes was conducted. Of the 160 undergraduates initially recruited to complete literature-based speaking tasks under either AI feedback or traditional teacher feedback conditions, 127 completed all three testing phases and were included in the final analysis. Oral fluency was assessed using speech rate (syllables per minute, SPM) and mean length of run (MLR) across pre-test, immediate post-test, and delayed post-test. Results showed that the AI group achieved significantly greater and sustained fluency gains compared with the teacher group. Cluster analysis revealed heterogeneous learner responses, with most students benefiting while a smaller subgroup exhibited limited progress. Cost-effectiveness analysis further demonstrated that AI feedback was nearly three times more efficient than teacher-only feedback in large-class contexts. The novelty of this study lies in integrating delayed testing, learner trajectory analysis, and economic evaluation within AI-mediated language learning. Findings have practical implications for educators and policymakers seeking scalable solutions to improve oral fluency in resource-constrained environments. The study contributes to the field by evidencing how lightweight AI can complement human instruction, offering both pedagogical and institutional value.

Zitieren

APA

Xue, C., Nitiasih, P. K., Padmadewi, N. N., & Mahayanti, N. W. S. (2025). Fostering EFL Oral Fluency through Lightweight AI Feedback in Literature-Based Speaking Tasks: A Quasi-Experimental Study in Chinese Higher Education. PPSDP International Journal of Education, 4(2), 1673-1690. https://doi.org/10.59175/pijed.v4i2.797

Vancouver

Xue C, Nitiasih PK, Padmadewi NN, Mahayanti NWS. Fostering EFL Oral Fluency through Lightweight AI Feedback in Literature-Based Speaking Tasks: A Quasi-Experimental Study in Chinese Higher Education. PPSDP International Journal of Education. 2025;4(2):1673-1690. doi:10.59175/pijed.v4i2.797

BibTeX
@article{Xue2025,
  author = {Xue, Cao and Nitiasih, Putu Kerti and Padmadewi, Ni Nyoman and Mahayanti, Ni Wayan Surya},
  title = {Fostering EFL Oral Fluency through Lightweight AI Feedback in Literature-Based Speaking Tasks: A Quasi-Experimental Study in Chinese Higher Education},
  journal = {PPSDP International Journal of Education},
  year = {2025},
  volume = {4},
  number = {2},
  pages = {1673--1690},
  doi = {10.59175/pijed.v4i2.797}
}
Kostenlos starten →