AI-generated corpus learning and EFL learners' learning of grammatical structures, lexical bundles, and willingness...

This research study examined the manner in which English as a Foreign Language (EFL) learners' willingness towards writing, grammatical construction, and lexical bundle acquisition were affected by language of AI-generated corpora. Eighty EFL students from China's Baotou Teachers' College participated in a repeated-measures quasi-experimental study and were split into controlled and trial groups. During the span of the 14-sessions, the experimental cohort underwent education that utilized AI-driven corpora, whereas the controlled cohort was instructed through traditional textbooks. Data were collected using pre- and post-treatment assessments, including tests on lexical bundles and grammar, as well as a willingness to write scale. The findings revealed that the trial group's performance was substantially improved than the controlled cohort regarding all outcomes. Specifically, trial group demonstrated higher mean scores for lexical bundles, grammar, and willingness to communicate. Statistical analysis, including mixed ANOVA, confirmed the significant effects of time and group membership on language proficiency and willingness to communicate. These findings suggest that the intervention positively influenced learners' language skills and attitudes toward writing. The interaction effects of time and group membership further highlighted the nuanced relationship between instructional interventions and learner outcomes. The study underscores the importance of integrating AI-driven language corpora into language teaching to enhance vocabulary, grammar, and communication skills. Pedagogical implications include the need for dynamic and engaging learning environments, while curriculum developers should consider incorporating data-driven learning approaches. Further research might identify optimal designs for learning and investigate its long-term impacts associated with such interventions. Policies should support equitable access to technology-enhanced language learning resources, promoting more effective language education practices overall.