Effects of AI-Generated Model Essays on Argumentative EFL Writing: A Quasi-Experimental Study
The New Educational Review · 2026 · 2026 (Vol. 83) · 197-207
DOI: 10.15804/tner.2026.83.1.13 ↗Kurz gefasst
Fachartikel in der Zeitschrift The New Educational Review, 2026. Verfasst von Saida Tobbi.
Abstract (Originalsprache)
This quasi-experimental pretest–posttest study tested whether guided engagement with AI-generated argumentative model essays plus explicit editing instruction produced larger gains in EFL argumentative writing than teacher handout exemplars. Ninety-six third-year undergraduates (two intact groups, n = 48 each) completed timed argumentative tasks before and after six weekly 90-minute sessions. Two assistant-lecturer raters scored essays blind to condition using an analytic rubric; objective indices (type– token ratio; errors/100 words) complemented ratings. Paired t tests showed within-group gains; an independent-samples t test on gain scores favored the experimental group (d = 0.72), with largest effects in lexical range and organization. Two experimental focus groups elucidated editing-to-transfer mechanisms.
Zitieren
APA
Tobbi, S. (2026). Effects of AI-Generated Model Essays on Argumentative EFL Writing: A Quasi-Experimental Study. The New Educational Review, 2026(Vol. 83), 197-207. https://doi.org/10.15804/tner.2026.83.1.13
Vancouver
Tobbi S. Effects of AI-Generated Model Essays on Argumentative EFL Writing: A Quasi-Experimental Study. The New Educational Review. 2026;2026(Vol. 83):197-207. doi:10.15804/tner.2026.83.1.13
BibTeX
@article{Tobbi2026,
author = {Tobbi, Saida},
title = {Effects of AI-Generated Model Essays on Argumentative EFL Writing: A Quasi-Experimental Study},
journal = {The New Educational Review},
year = {2026},
volume = {2026},
number = {Vol. 83},
pages = {197--207},
doi = {10.15804/tner.2026.83.1.13}
}
EUVIMED